assisted reproductive technology research topics

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Review Article
• Open access
• Published: 01 March 2024

The prospect of artificial intelligence to personalize assisted reproductive technology

• Simon Hanassab   ORCID: orcid.org/0000-0001-9112-0486 1 , 2 , 3   na1 ,
• Ali Abbara 1 , 4   na1 ,
• Arthur C. Yeung 1 , 4 ,
• Margaritis Voliotis 5 , 6 , 7 ,
• Krasimira Tsaneva-Atanasova   ORCID: orcid.org/0000-0002-6294-7051 5 , 6 , 7 ,
• Tom W. Kelsey   ORCID: orcid.org/0000-0002-8091-1458 8 ,
• Geoffrey H. Trew 1 , 9 ,
• Scott M. Nelson 9 , 10 , 11 ,
• Thomas Heinis 2   na2 &
• Waljit S. Dhillo   ORCID: orcid.org/0000-0001-5950-4316 1 , 4   na2  

npj Digital Medicine volume  7 , Article number:  55 ( 2024 ) Cite this article

12k Accesses

10 Citations

18 Altmetric

Metrics details

• Endocrine reproductive disorders
• Infertility
• Machine learning

Infertility affects 1-in-6 couples, with repeated intensive cycles of assisted reproductive technology (ART) required by many to achieve a desired live birth. In ART, typically, clinicians and laboratory staff consider patient characteristics, previous treatment responses, and ongoing monitoring to determine treatment decisions. However, the reproducibility, weighting, and interpretation of these characteristics are contentious, and highly operator-dependent, resulting in considerable reliance on clinical experience. Artificial intelligence (AI) is ideally suited to handle, process, and analyze large, dynamic, temporal datasets with multiple intermediary outcomes that are generated during an ART cycle. Here, we review how AI has demonstrated potential for optimization and personalization of key steps in a reproducible manner, including: drug selection and dosing, cycle monitoring, induction of oocyte maturation, and selection of the most competent gametes and embryos, to improve the overall efficacy and safety of ART.

Similar content being viewed by others

Optimizing oocyte yield utilizing a machine learning model for dose and trigger decisions, a multi-center, prospective study

Machine learning for sperm selection

Pseudo contrastive labeling for predicting IVF embryo developmental potential

Introduction.

Since the birth of the first baby conceived through in vitro fertilization (IVF) in 1978, the development of assisted reproductive technology (ART) has evolved significantly. Over the last 40 years, ART has provided infertile couples with the possibility to conceive, culminating in the birth of over eight million children 1 . IVF protocols are complex and require intensive monitoring, with clinicians and embryologists responsible for several key decision points prior to and during the cycle (Fig. 1 ). Although several of these decisions have a solid evidence base, many are highly subjective and will vary immensely based on clinical experience with an inevitable non-reproducible impact on clinical outcomes—leading to the mantra that ART is an art.

Potential targets for the application of artificial intelligence and machine learning methods during clinical and embryological steps in assisted reproductive technology (ART). Investigations of infertility and pre-treatment counseling are not captured here and discussed independently in Section Pre-treatment counseling. The order and timings of the steps can differ depending on the ART protocol used. Figure created with BioRender.com.

Given these limitations, there is increasing recognition that alternative data-driven approaches that harness the large number of ART cycles undertaken and facilitate objective, consistent, and optimal decision-making may be associated with improved outcomes. Large amounts of data generated during IVF cycles have enabled interdisciplinary researchers to propose artificial intelligence (AI) methodologies to drive individualized approaches. These have ranged from algorithmic drug dosing tools, to ‘human-in-the-loop’ AI clinical decision support systems (CDSSs) for embryo selection, whereby humans are supported by AI but ultimately make the final decision. Harnessing the symbiosis between the experience of clinicians, and personalized recommendations from AI models based on the one million cycles undertaken annually, has the potential to synergistically improve clinical outcomes. In this review, we examine current implementations of AI models within ART, and future prospects concerning their utility, efficacy, and application in the field.

Artificial intelligence methods for assisted reproductive technology

AI is an overarching term that encompasses a growing number of subfields including machine learning (ML), robotics, and computer vision (Fig. 2 ). Principally, ML methods can learn patterns from data and draw inferences, and therefore build models that optimize/personalize ART protocols for a specified outcome. Traditionally, ML can be under either a supervised, unsupervised, or reinforcement learning framework. In supervised learning, data are labeled as inputs and outputs with the goal being to develop models that capture the relationship between the two, which can be used to predict outputs when presented with new, unseen inputs. Conversely, in unsupervised learning, models are built to capture the structure (e.g., clustering) of data with no output labels (‘unlabeled’) that can be used to interpret new, or generate synthetic data. Reinforcement learning trains an ML agent that interacts with a defined environment towards achieving a goal and receives a ‘reward’ for its actions.

A Venn diagram providing a holistic view of the artificial intelligence (AI) landscape, with a particular focus on machine learning (ML) methods. ML is a subfield that is often used in conjunction with other AI subfields, such as computer vision. Some methods can be used in alternative learning frameworks however their most common current manifestations are presented here.

Supervised methods include decision trees, linear/logistic regression, k -nearest neighbors, support vector machines, random forests, artificial neural networks (ANNs), and more. Decision trees are models used to classify or predict outcomes based on input data. They can effectively capture non-linear relationships and can be visualized intuitively as tree-like structures: starting from the root, each branch represents a decision rule to select which subsequent branch should be followed; the final nodes (‘leaves’) of the tree represent outcomes. Extending this to an ‘ensemble’ of trees inspires the random forest algorithm, where each tree is trained on a random partition of the data and its input features. The final prediction is determined by a voting mechanism, combining the predictive power of all decision trees. This generally makes the model less prone to ‘overfitting’, a phenomenon whereby a model may perform very well on training data but poorly on new, unseen data. Supervised methods have widespread applications with tabular (i.e., numerical or categorical) outcomes in ART.

ANNs are networks of connected computational units representing artificial neurons—they receive inputs, process them, and signal the result to other neurons connected to them, in a multi-layered structure (e.g., the multi-layer perceptron algorithm). The input layer receives data to be processed, and the output layer presents the model output. The strengths (‘weights’) of connections between artificial neurons comprise the parameters of the ANN and are calibrated during model training.‘Deep’ learning is ANNs with complex architectures comprising many layers, an example being convolutional neural networks (CNNs), useful for spatial, grid-like data (e.g., embryo images).

As for unsupervised methods, k -means is a popular algorithm for clustering data into k -groups based on the distance of data points from the centroid of each group. Another example is generative adversarial networks (GANs), where one network is trained to generate synthetic data whilst the other discriminates synthetic from real data. The two networks are trained in parallel, competing as adversaries, resulting in better discrimination between synthetic and real data. Multimodal generative AI has recently caught mass media attention, especially through both text (e.g., ChatGPT and Med-PaLM) and text-to-image generators (e.g., DALL-E), which have been evolving rapidly in performance since their inception 2 . These frameworks bring together large language models (LLMs), a type of natural language processing built with ANNs, and diffusion models, an alternative generative methodology to GANs based on iterative de-noising to estimate how image data are distributed to therefore generate a desired image 3 .

During model development, it is standard practice to use ‘training’, ‘validation’, and ‘test’ datasets: ‘training’ to fit the model, ‘validation’ to fine-tune the model’s hyperparameters, and a ‘test’ set to independently evaluate the model’s performance. For generally more reliable estimates of model performance, cross-validation can be used to evaluate the model on multiple training/validation data splits. Using test datasets that are externally unseen and temporally different (e.g., from a different clinic) can provide further reassurance of a model’s generalizability. The fundamental choice of ML algorithm for a certain task is multifaceted and often driven by contextual reasoning. Nevertheless, Table 1 presents some rules-of-thumb regarding popular ML methods (Fig. 2 ) within the context of ART.

Pre-treatment counseling

Classically, age-stratified population estimates have been used to inform patients of their overall chance of success, however, these often fail to incorporate important determinants of outcome such as previous treatment cycle attempts or for treatment-naive patients their ovarian reserve and likely ovarian response. To try to tailor these models further both population data and clinic-specific datasets have been used to develop models for a variety of outcomes including for cumulative live births across multiple cycles 4 . These models are now being used by both patients and a range of stakeholders to manage access to care (national healthcare services, insurance providers) and clinics, or third-party companies offering shared-risk financial programs 5 . Moreover, the emergence of AI chatbots using LLMs could improve efficiency in the initial assessment of infertility. A recent ‘Fast Track to Fertility’ program using semi-automated two-way text messages reduced the time to complete a workup by 50% 6 . The deployment of LLMs for fertility assessment offers unique challenges and currently remains experimental, whilst the frameworks for validation and regulation of such systems are yet to be formalized 2 , 7 .

Gonadotropin dosing for ovarian stimulation

Ovarian stimulation (OS) is used to stimulate the growth of multiple ovarian follicles in order to result in multiple oocytes for retrieval 8 . IVF treatment is a profligate process as not all follicles yield oocytes, not all oocytes will fertilize, and not all embryos will develop, implant, or be capable of becoming healthy babies. Various preparations of gonadotropins exist but most will contain a supra-physiological amount of follicle-stimulating hormone (FSH) to extend the ‘FSH-window’ by maintaining high FSH levels, and induce multi-follicular growth 9 . Optimization of the gonadotropin dosing regimen can maximize the number of follicles with respect to ovarian potential 9 . As such, an optimal initial dose of FSH can ensure sufficient follicles are recruited, whilst avoiding the recruitment of too many follicles (often defined as >15 oocytes at pickup), and an increased risk of ovarian hyperstimulation syndrome (OHSS) 8 , 10 .

The application of ML approaches to retrospective datasets for model learning has demonstrated the potential to personalize FSH dose as summarized in Table 2 . Fanton et al. aimed to identify the 100 most similar patient profiles to each patient, to then generate individualized dose-response curves 11 . The authors reported limitations including a protocol-agnostic approach, and that 87% of cycles included both pure FSH and Menopur (for luteinizing hormone (LH)-like activity) during OS 11 . The methodology was further evaluated against the national US database (SART CORS) including 365,473 patients and reported upon in conference proceedings 12 . The results similarly predicted that an increased number of two-pronuclear fertilized embryos (2PNs) and blastocysts could be retrieved whilst using significantly lower total FSH doses, key in reducing high medication costs for patients 12 . Nevertheless, OS protocols vary across clinical practice, and the generated dose-response curves presented less confidence in predicting oocytes with lower doses of FSH administration 11 , which are the norm in Europe (where 150-225 IU is suggested for normal responders 8 ). Therefore, it is necessary to determine whether the proposed models are directed at certain geographies or intend to be universal. Setting a precedent for the conduct of future multi-center studies is central to achieving either objective—Ferrand et al. successfully leveraged a federated learning framework 13 , a potentially effective approach that allows data to be kept decentralized and private, whilst deploying ML models for collaborative training between clinics 14 , 15 .

Recent studies have also focused on the effects of demographic, endocrine, and genetic data to optimize OS, and therewith predict the retrieval of mature oocytes 16 , 17 , 18 . Although these are retrospective studies, they highlight the need to explore available characteristics and further assess their impact on clinical outcomes when determining dosing regimens, whereby endocrine monitoring or genomic sequencing for ART cycles may be efficacious for some patients 19 , 20 . To best identify such predictors in an unbiased manner, the treatment cycles of patients should not exist in both the training and test sets 21 . An independent test set of patients should be partitioned at random, or if cross-validation is employed, cycles from the same patient must not exist across the training and test folds.

Ultimately, determining the efficacy of introducing individualized gonadotropin dosing algorithms into the clinic will require appropriate validation across different geographies. The three prospective international multi-center randomized controlled trials (RCTs) for follitropin delta (recombinant-FSH; Ferring Pharmaceuticals) that assess a unique algorithm to facilitate individualization of dose based on anti-Müllerian hormone (AMH) and body weight are an apt example of that critical step 22 , 23 , 24 . The retrospective studies in Table 2 would benefit from similar prospective validation in multiple centers to establish whether their adoption in the clinic is appropriate and of value for patients.

Induction of oocyte maturation

Once multiple follicles have grown during OS, a hormonal trigger is administered to mature oocytes in preparation for retrieval. The triggering agent is most efficacious when follicles are neither too large nor too small 10 . In turn, AI/ML techniques have been harnessed to optimize the trigger day (TD) as summarized in Table 3 . Our research team previously developed a random forest model to determine follicle sizes on TD that most contributed to the number of mature oocytes retrieved 25 . Maximizing the number of follicles sized 12-19 mm on TD was determined as optimal for yielding mature oocytes and could be used as a feature in conjunction with baseline endocrine characteristics to predict oocyte yield 19 .

A more recent study leveraged patients that had ultrasound scans both on the day before trigger, and on the true TD, to learn why a clinician might decide to wait a further day to trigger 26 . They found follicles sized 16-20 mm as most contributory in determining optimal TD, and predicted superior outcomes in terms of 2PN and blastocyst yield compared solely to a clinician’s decision 26 . With a similar methodology but using a simpler model, Fanton et al. confirmed the findings with even further granularity and showed follicles sized 14-15 mm were most predictive on TD, whilst those sized 11-13 mm on the day prior to triggering were most contributory 27 . The aforementioned studies employed ML methods which show predictor importance measures against the desired outcome (oocytes retrieved), and therefore provide a useful data-driven target for oocyte maturation based upon many previous IVF cycles 25 , 26 , 27 . Transparent models such as these should be favored at embryonic stages of AI-driven developments, to ensure clinicians and patients can gain trust towards CDSSs as part of ART workflows 28 , 29 . It is crucial to take into account the nuances of workload management in daily clinical practice in order to incorporate AI models into workflows effectively 30 . Real-world data where ultrasound scans may not be conducted every day can challenge the precision of models developed to assess TD or misrepresent the predictive capacity of certain features.

A proof-of-concept CDSS by Letterie and MacDonald (Table 2) also considered a decision point to trigger or cancel the cycle 30 . This notion was further developed in a later study looking specifically at TD assignment to optimize the retrieval of oocytes 31 . Features included pre-cycle characteristics, as well as estradiol level and follicle diameters determined on the single ‘best day’ for assessing TD, for which baseline AMH alone was most predictive 31 . A stacking model was trained, which compounds the predictive power of multiple ML models to improve overall robustness. This CDSS fulfills the need for streamlining follicular monitoring that may arise from reasons such as long-distance travel to clinics or unprecedented public health constraints. In response to the constraints enforced by COVID-19, Roberston et al. demonstrated that day-5 of OS would be the ‘best day’ for predicting both the risk of OHSS and optimal TD 32 . Both these studies highlight reducing monitoring in certain clinical settings may be possible, which could reduce resource requirements in the clinic, and the burden upon patients. The timing of the TD is a multifaceted decision point and therefore to confirm utility in practice, prospective validation of the developed models in diverse populations would be a prudent next step forward.

In the embryology laboratory

The application of AI in the embryology lab has attracted significant recognition in recent years and has been reviewed comprehensively 33 , 34 , with more recent developments summarized here (Tables 4 , 5 , and 6 ). The capacity of AI techniques to analyze large amounts of complex data such as images and time-lapse objectively, whereby non-invasive assessment of gametes and embryos can be done in real-time, has significant potential for future impact in achieving healthy live birth. This can lessen the need for specialist embryology resources whilst automating some of the processes involved to reduce costs.

Sperm assessment

Computer-aided sperm analysis.

Standard semen analysis comprising of concentration, motility, and morphology assessment remains the first-line investigation of pre-treatment male fertility potential. Computer-aided sperm analyzers (CASA) aim to reduce intra-operator subjectivity and variability associated with manual assessment while standardizing and increasing throughput capacity. CASA analysis of sperm concentration and motility have shown a good correlation with manual assessment 35 , while estimates of progressive motility are also significantly linked to both in vivo and in vitro fertilization rates 36 , 37 , 38 , 39 . However, CASA-based morphological assessment tends to correlate the least with manual assessment, likely as a result of heterogeneity within a given semen sample and the subjective nature of interpretation 35 .

The latest WHO manual on sperm analysis 40 (2021) recognized the ability of CASA to accurately determine sperm concentration and progressive motility parameters through the use of fluorescent DNA stains and tail-detection algorithms 41 . These advancements have improved the distinction between immotile spermatozoa and particulate debris; a problem that has led to the overestimation of concentration, and underestimation of progressive motility, since the inception of computer-aided systems.

At a population level, ML algorithms could be a useful to identify individuals at risk of an abnormal semen profile. An ANN based on an 11-question demographic characteristic questionnaire (including age, alcohol consumption, smoking status, urbanization and occupational exposures) achieved 92.9% accuracy in predicting abnormal sperm concentration, and 85.7% for predicting any sperm abnormality 42 . Although only developed in a small cohort of 141 men, if replicated, an AI-driven triage model could be used as a preliminary screening tool with early recourse to diagnostic testing.

Further, an ANN using semen parameters as inputs in 177 men was able to predict seminal plasma biochemical markers including fructose, zinc, and total protein content 43 . The added value of these biochemical parameters over standard semen analyses is still unclear, but a number of omics-based markers in seminal fluid have been identified as helpful in determining fertilization prognosis in a cost-effective manner 44 . Incorporating these techniques into the IVF clinic is challenging, namely due to initial set up costs and specialized techniques required for analysis. Moreover, whether these markers and profiles could drive selection of an individual spermatozoon for fertilization remains unclear.

Accurate assessment of sperm motility is paramount in fully understanding genetic and biochemical factors that may impact normal fertilization and thus plays a key role in selection for ART. Motility prediction based on deep learning using sperm videos has been examined with promising results 45 , 46 , 47 . AI software may begin to allow correlation of kinetic motility patterns with other crucial factors such as sperm morphology, likelihood of fertilization, or blastocyst formation to aid in selection for intra-cytoplasmic sperm injection (ICSI) in real-time 48 , 49 . These studies show the potential of incorporating temporal features into deep learning models to extract insights into sperm motility consistently and efficiently.

Staining of spermatozoa is currently required to identify morphological abnormalities and defects for diagnostic purposes. However, given that the staining of sperm affects their vitality and motility, tested spermatozoa are no longer viable for use in ICSI and thus, do not aid in sperm selection for fertilization 50 . Consequently, morphological assessment of a single spermatozoon in a non-invasive manner using AI techniques is of interest for sperm selection 34 . Some models consider specifically the sperm head morphology 51 , 52 , 53 , 54 , whereas others consider a more comprehensive analysis of the whole sperm 55 .

WHO describe eleven different sperm head abnormalities by taking into account shape, size, and consistency 40 . Some of these subtypes present further challenges, with their morphology forming a vast continuum with overlaps, such that discrimination is complex to the naked eye. Using a dictionary learning approach combined with segmented microscopic sperm head images, Shaker et al. achieved a 92.3% accuracy in distinguishing between four sub-types against a ground truth dataset agreed by three experts 52 .

Open datasets of spermatozoa are becoming accessible to researchers and have been used to benchmark different models against one another 51 , 52 , 56 . Latest deep learning advancements with CNNs are capable of detecting morphological deformities in spermatozoa head, acrosome, and vacuole in real-time using low-magnification microscopes (400-600x) without staining and with increased objectivity 56 , 57 .

Non-invasive AI methods are also capable of assessing morphological features of immotile or frozen sperm that are difficult to characterize manually. Current viability tests require cytotoxic staining that renders individual spermatozoon unusable for ICSI. Recently, Jiang et al. described an AI model capable of identifying viable sperm based on a single bright-field image without the need for any sample processing or reagents 58 . The model exhibited 94.9% accuracy, 97.0% sensitivity, and 93.3% specificity, based on subtle morphological changes to the cell nucleus. Incorporation of such AI models into existing CASA systems could further reduce the need for sperm staining in the future, especially in the context of surgically retrieved or frozen sperm with unknown viability.

To our knowledge, no computer-aided systems exist to improve the surgical retrieval of sperm yet. Current testicular sperm extraction techniques for ICSI can be challenging, with outcomes being greatly operator-dependent 59 . However, AI techniques to aid identification of sperm from biopsies during testicular sperm extraction have been investigated. Wu et al. describe a deep CNN capable of finding sperm in testicular biopsy samples with good accuracy (mean average precision of 0.74) but did not compare this to standard embryology techniques 60 . ML models employing 16 preoperative assessment variables (e.g., hormonal parameters, genetic, demographic, lifestyle, and urogenital history) have also been shown with moderate performance to predict the success of testicular sperm extraction 61 . Given the clinical implications of not pursuing surgical sperm retrieval (i.e., unequivocal use of donor sperm), further external validation of this promising model is required. The inclusion of additional biomarkers such as more detailed genetic information, seminal plasma microRNA, or additional hormones, as a way of further improving model performance, would also be of interest.

Sperm selection for ICSI is not standardized and WHO guidelines are interpreted subjectively by embryologists. High-throughput AI models have the potential to be more objective and tackle the fundamental challenge of selecting individual sperm with the best potential for embryo formation from a sample of over 10 8 gametes 50 . Nonetheless, with respect to morphology, there are currently no studies that assess AI performance against manual assessment according to WHO guidelines 34 . Indeed, the potential performance of AI networks is directly linked to the quality of the database used for training, as well as the caliber of data used as input. Progress on its use in sperm selection would benefit from global collaboration between clinical and laboratory teams to build a robust and definitive database of sperm images to establish a consensus ground truth.

DNA fragmentation

Existing techniques for sperm DNA fragmentation similarly lack data at the single spermatozoon level. Modern-day tests of DNA integrity are invasive and conducted at the sample level, making them an unsuitable metric in the selection of individual sperm for ICSI. McCallum et al. described a CNN trained using a set of 1064 images of individual sperm cells of known DNA integrity to provide a DNA integrity prediction from a single bright-field image in under 10 ms 62 . Recently, Kuroda et al. described further progress with their AI-augmented sperm chromatin dispersion (SCD) test kit capable of assessing DNA fragmentation in >5000 spermatozoa at once, compared to a limited 300 in the widely commercially-used Halosperm SCD test 63 . The improved kit showed a good correlation with the conventional test that requires manual counting (Halosperm G2; r  = 0.69, p  = 0.02). DNA fragmentation counting took 5 min. in the automated device compared to around 20 min. with the manual method 63 .

Emerging evidence increasingly suggests that sperm DNA fragmentation is associated with reduced male reproductive capability and can be assessed in combination with conventional sperm analysis 64 . However, routine testing remains contentious and may not necessarily provide predictive value 65 . Other technical limitations exist, in particular the use of different staining, microscopes, and assays for DNA fragmentation that can challenge the training of an accurate AI model. Guidelines for testing, and optimal techniques for testing sperm DNA fragmentation have been proposed 66 , 67 , but testing is still not widely recommended. Progress in this field thus relies on the standardization and optimization of DNA fragmentation assays, prospective evaluation of its impact on ART outcomes, and the development of therapies to improve sperm DNA fragmentation levels 68 . Should this be achieved, ML algorithms that can combine morphological, motility, and DNA fragmentation data with outcomes such as fertilization, miscarriage, and live birth rates, could standardize, and vastly improve, single sperm assessment/selection by reducing the subjective and inter-variable outcomes between embryologists.

Oocyte assessment

Nuclear maturity of human oocytes can only be verified by observation of the extruded polar body, which requires removal of the cumulus 10 . Automated, non-invasive methods to assess nuclear and cytoplasmic maturity and future reproductive potential would be desirable, particularly for fertility preservation. Accurate prediction of oocyte quality and fertilization prospects would allow better estimation of personalized live birth predictions from a pool of cryopreserved oocytes. Consideration of whether this is sufficient to realize a desired family size may dictate the need for further cycles of OS and cryopreservation. Clinicians would also be able to manage expectations for success and reduce the number of poor-quality embryos with low implantation potential 69 .

Currently, assessment of nuclear oocyte maturity is performed visually by embryologists in a subjective manner prior to fertilization. Oocyte scoring systems assessing cytoplasmic morphological features such as the presence of vacuoles, degree of perivitelline space, and cytoplasmic granularity, among others, have long been proposed as predictors of insemination outcome but remain points of contention as prognostic indicators of embryo development and implantation 70 , 71 . Substantial labeled datasets of oocytes are scarce—as such, Kanakasabapathy et al. combined a retrospective dataset of oocyte images with known fertilization outcomes alongside synthetic oocyte images generated by a GAN to form a synthetic CNN 72 . This synthetically-extended CNN outperformed the raw CNN, and delivered an accuracy of 82.58% with an AUC of 0.81 in identifying oocytes that would fertilize normally to form two-pronuclear zygotes (2PNs), versus those that would not (non-2PNs) 72 . This study showed the value of using AI to augment the training, predictive power, and robustness of existing CNNs available for the embryology lab, perhaps widening their scope of use in ART 73 .

A non-invasive CNN-based software, VIOLET™ (Future Fertility), has been shown to predict fertilization and blastulation with 91.2% and 63% accuracy respectively, based on morphological features of 2D oocyte images. The tool’s performance was much quicker and also outperformed expert embryologists in accuracy 74 . VIOLET™ aims to give users undergoing oocyte cryopreservation a personalized estimate of live birth potential based on the morphology of oocytes cryopreserved as opposed to generalized age-related outcomes. Similarly, the MAGENTA™ tool employs 2D images of denuded oocytes and a similar morphology-based CNN to score oocytes and predict the potential for high-quality blastocyst formation with good accuracy 75 . Though promising in correlating oocyte morphology with blastocyst potential, their estimates lack interplay with potential male factor subfertility and could benefit from the incorporation of clinical variables such as BMI or endometriosis, to enhance the prediction of outcomes such as clinical pregnancy or live birth.

More recently, a non-invasive gene expression test was prospectively trialed by Link et al. 76 . The ‘OsteraTest’ software is composed of eight ML modules and uses a 25-gene network to predict oocyte quality based on cumulus cells 76 . This bioinformatics-inspired approach was able to non-invasively predict oocyte development to a day-5 blastocyst with 86% accuracy 76 . Though further large-scale validation is necessary, this type of AI approach could change current practices in oocyte selection prior to cryopreservation and ICSI, as well as reduce the pool of embryos formed, cryopreserved, and tested, prior to embryo transfer. This may be particularly beneficial in countries with regulatory frameworks surrounding embryos such as Poland, where only six oocytes may be fertilized per cycle, or Germany where no more than three embryos can be stored per treatment attempt. Additionally, it may guide egg sharing or donor oocyte cycles and inform on how to distribute oocytes evenly or the total within a cohort depending on blastocyst potential.

Although these approaches provide direction for further research, the data must be viewed with caution until published in peer-reviewed journals. In developing an AI model, it is imperative to define a set end goal such as oocyte quality following oocyte cryopreservation. If fertilization is planned and blastocyst potential is being predicted, then spermatozoon quality and other male confounders should be considered. Proposed biomarkers to predict oocyte potential include follicular fluid markers (insulin-like growth factor, zinc levels 77 ), cumulus-oocyte complex composition 78 , and cytoplasmic features like mitochondrial function 79 . Consideration of these methods to guide oocyte selection in the future would also require analysis into whether they are feasible in daily practice or in fact as cost-effective as fertilizing all suitable oocytes 80 .

Embryo assessment

Embryo selection based on morphological assessment is an important predictor of success in IVF cycles but is primarily based on static visual observations at specific developmental time points. Information obtained in this manner is not only highly subjective with great inter-operator variability but also diminishes the dynamic nature of a developing embryo in culture, thus limiting its accuracy. AI-driven embryo analysis is suited to predicting developmental potential, non-invasive aneuploidy assessment, and ultimately the selection of an embryo with the best live birth potential for transfer.

Morphokinetics and morphology

Examples of developments in embryo evaluation include the assessment of pronuclear stage embryos to differentiate between 2PN and non-2PN zygotes 81 , 82 . Morphokinetic data such as cytoplasmic movements have also shown potential to predict blastocyst formation at early cleavage stages in a time series-based ANN model 83 . Further assessments of interest include morphological classification of pronuclei size and arrangement to monitor embryo development 84 . CNN models showed comparable results to manual labeling, albeit with high precision and reproducibility at a fraction of the time required by clinicians (12.18 s vs. 130 hrs.) 84 . Despite promising results, the standard morphological assessment remains the international consensus which is subjective and labor-intensive.

Time-lapse images combined with automated morphology assessment of embryos based on CNNs have shown promise, capable of outperforming individual embryologists with excellent accuracy 85 , 86 . Other fully automated deep learning-based models using time-lapse images such as iDAScore (Vitrolife) have shown the ability to accurately assess embryo morphology without the need for concurrent embryologist assessment or annotation, and predict implantation outcome 87 , 88 , 89 . The benefit of using time-lapse incubation systems and/or AI technology in the embryo selection process is yet to be proven as superior to current means in double-blind RCTs 90 , 91 . The SelecTIMO trial recently showed no improvement in cumulative live birth rates when using uninterrupted culture conditions with routine morphological embryo selection compared to a time-lapse based embryo selection algorithm alongside uninterrupted culture for day-3 embryos 92 . With no improvement in cumulative pregnancy rates or time-to-pregnancy, it may be that the time-lapse selection method may not improve pregnancy rates, however, whether this applies to day-5 embryos is still to be clarified. Nevertheless, the time-lapse technology was not inferior and therefore could achieve similar outcomes in an automated and less subjective manner. Importantly, with modern advancements in cryopreservation, it is likely that the most viable embryos will eventually be transferred if needed. Additionally, human input may be needed to aid the assessment of embryo quality, for example, by repositioning embryos to get a better view, which should be taken into account when considering the application of an AI for this task. Validation data from the VISA Study (ClinicalTrials.gov Identifier: NCT04969822), a noninferiority, prospective, multi-center RCT may further reflect the clinical impact of AI-driven systems compared to manual morphology assessment by embryologists for day-5 embryos. Such studies highlight the necessity for the accuracy of predictions made via AI techniques to be prospectively validated prior to adoption into clinical practice with appropriate mitigation of study biases and evaluation of cost-effectiveness 20 , 93 .

Recently, a biomarker-scoring CDSS based on 799 blastocyst videos, CHLOE EQ ™ (Fairtility), has been described and takes into account patient and embryo data including blastocyst diameter, degree, and time of expansion, and other morphokinetic markers. Though preliminary results are promising, these new systems still require external validation and larger-scale prospective studies before widespread adoption to realize the end goal of fully automated blastocyst assessment and accurate embryo prognosis 94 , 95 . It is paramount that future algorithms focus not only on the competitive selection of the best embryos for culture and transfer but also can differentiate between embryos that are otherwise morphologically indistinguishable to the naked eye, wherein the real challenge lies.

Rates of pre-implantation genetic testing for aneuploidy (PGT-A) as a screening tool to improve clinical outcomes in ART cycles have increased in recent years. Currently, PGT-A is performed by trophectoderm biopsy on blastocysts followed by whole-genome or targeted DNA amplification and a next-generation sequencing assay. Multiple blinded non-selection studies have now shown a high prognostic failure of live birth when an aneuploid result is obtained 96 , 97 . Furthermore, discarding uniformly aneuploid embryos is unlikely to have a meaningful impact on cumulative live birth rates, especially in women over 35 years of age where it is more likely to be employed 98 . As modern invasive techniques still bring technical and financial challenges, non-invasive AI-driven PGT-A could offer the benefits of PGT-A without embryo manipulation and biopsy. Recent single-center studies have shown ongoing validation of AI models feeding time-lapse imaging data into CNNs to predict ploidy status from abnormal morphokinetic patterns with good accuracy 99 , 100 . These models may not replace PGT-A but highlight the potential for PGT-A triage and well-informed guidance towards embryo selection in a non-invasive manner 99 , 100 , 101 , 102 . Once again, further validation and large multi-center datasets must be compiled for standardization and generalization of these AI-driven models.

A comprehensive understanding of the embryo at a molecular level may present another adjunct for the high throughput and comprehensive capabilities of AI-driven predictive models in the future. Various metabolomic signatures of an embryo have been investigated over the years, mainly pertaining to metabolites or biomarkers in spent culture media as a reflection of complex physiological and pathological responses and in turn, reproductive potential or ploidy status. Conflicting results to this approach have been shown 103 , 104 , 105 , 106 , 107 , while a previous meta-analysis including four RCTs and a total of 924 women showed no meaningful effects for metabolomic assessment on clinical outcomes 108 . Interestingly, an ANN employing a combination of conventional embryological data and thirteen nuclear magnetic resonance spectroscopy-identified metabolite levels has shown promise in predicting blastocyst implantation, though at a very small scale with a test dataset of twelve spent culture media 109 .

Current limitations of the omics approach lie within the vast variability in culture media components used and handling of spent media, contrasting infertility phenotypes, definitive biomarkers predictive of reproductive potential, and a general lack of conclusive evidence that fertility outcomes can be optimized through omics profiling. Though non-invasive, highly specific, and perhaps crucial towards a better understanding of gamete development, it is unclear whether omics profiling can effectively contribute to an improvement in clinical outcomes or will remain principally a research tool 110 . Furthermore, the complexities of omics analysis and interpretation of output data present significant barriers to adoption in daily laboratory practice.

Embryo quality aside, reproductive outcomes also depend on implantation and the endometrium. The construction of models should also integrate features of the uterus and crosstalk between an embryo and the endometrium. To date, the clinical benefit of an endometrial receptivity array (ERA) for assessment has yet to be proven 111 . The invasive nature of biopsy for endometrial receptivity testing, the time needed for results preventing immediate embryo transfer, and the potential accuracy of the diagnostic test itself are further limitations 112 . AI is however well suited to drive collaboration between ART clinics and omics-focused research groups, on account of its ability to perform large-scale data throughput and analysis. Whether these approaches will alter conventional therapies remains unclear, particularly as diagnoses such as true recurrent implantation failure and its relevance are being hotly debated currently 113 . However, given the lessons to date, the value of any ‘AI-omics’ platform should be validated in appropriately powered RCTs.

Conclusions and future prospects

With respect to ART, several groups have developed CDSS frameworks or decision-making tools for use at key decision-points in the clinic, and/or embryology laboratory 17 , 30 , 31 . Personalization in further avenues could better improve the clinical outcomes of ART. Ovarian response has been shown to vary significantly depending on ovarian reserve, between ethnic groups 114 , 115 , FSH receptor genetic polymorphisms 116 , and body weight 19 , 117 . Therefore, incorporating such factors which influence pharmacokinetic parameters when dosing gonadotropins 9 , 19 , 20 , or suppressing premature ovulation 20 , 118 , may be beneficial. ML methods could also help tailor luteal phase support regimens to certain patient subgroups, where a lack of clinical consensus currently exists 119 .

The ubiquity of electronic health records (EHRs) has accelerated the development of CDSSs 15 . A predominant barrier to adoption is trustworthiness, especially with ‘black-box’ AI systems 29 , which has led to transparency being a key characteristic preferred by clinicians as such models offer simpler interpretations, although may compromise accuracy when applied to more complicated learning tasks 28 . Implementations of ‘black-box’ models are evolving, especially for embryological analyses, due to the data being primarily image-based; in turn, efforts in explainability have emerged to seek insights for model generalizability, fairness, and trustworthiness 94 , 95 . Misleading conclusions may be reached if clinical inference is neglected during the decision-making process since such methods are often correlation-based and prone to ‘overfitting’ 120 . Generating counterfactual examples in this context, such as: “what if the optimal TD was yesterday(?)", or “what if the other embryo were implanted(?)", are generally unavailable—and to further exacerbate this—ground truths are often based upon clinical guidelines/scoring rather than objective outcome labels. The emergence of omics analyses offers an alternative, and arguably more efficient, solution for clinical and embryological assessment, although advancements currently remain of a preliminary nature 18 , 108 . Ultimately, appropriate assessment of CDSSs for ART is necessary in practical, ethical, and clinical contexts prior to clinical adoption. Rigorous validation with comprehensive standardized reporting is essential for establishing trustworthy models before attempting viable integration into clinical workflows 21 , 121 . Research conduct and reporting guidelines such as PROBAST-AI are in progress for the wider field of AI for healthcare, and with this at hand, a more granular and contextual guideline for AI in the domain of ART can be proposed 122 , 123 .

Salient efforts from both academia and industry have validated the utility of retrospective data to enable data-driven decision-making for ART 123 . To ensure viable deployment, these models can benefit from larger, multi-center datasets that incorporate both heterogeneous patient populations and also capture the idiosyncratic nature of clinical practice worldwide. Achieving this is best achieved through a collaborative effort from all stakeholders representing multiple disciplines across the AI and healthcare landscape 21 . Furthermore, streamlining workloads is an essential objective of CDSSs, and seamless implementation with, or within, EHR systems are essential to not inadvertently decrease the efficiency of clinical workflows. Prospective validation (e.g., well-designed RCTs) with relevant outcome measures is a key step to assess the efficacy and efficiency of these models in clinical environments and thus demonstrate impact on patient outcomes. With such efforts in place, a comprehensive end-to-end CDSS seems a plausible future goal. Whether this paradigm should extend to an autonomous AI clinician within the ART domain remains an open and contentious question. The use of AI to automate some of the tasks currently performed by clinicians or laboratory staff could have implications in training and a potential loss of expertise in the workforce, but may also free up staff time to focus on more challenging and physically demanding technical processes. Reflections on the current literature to date elicit valuable questions regarding future studies, including determining the specification of what should be measured/captured, to what precision, and how often. Decision points cannot necessarily be considered in isolation, and the relationships between some of the key topics described in this review require further interdisciplinary research to prioritize the individualization and utility of certain decisions over others. The intersection of AI and ART undoubtedly remains a nascent and valuable field of study, which has the potential to reduce intensive resources, whilst ultimately improving clinical outcomes for patients.

Data availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Fauser, B. C. Towards the global coverage of a unified registry of IVF outcomes. Reprod. Biomed. Online 38 , 133–137 (2019).

Article   PubMed   Google Scholar  

Thirunavukarasu, A. J. et al. Large language models in medicine. Nat. Med. 29 , 1930–1940 (2023).

Article   CAS   PubMed   Google Scholar  

Gu, S. et al. Vector quantized diffusion model for text-to-image synthesis. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition 10696–10706 (IEEE, 2022).

McLernon, D. J. & Bhattacharya, S. Quality of clinical prediction models in in vitro fertilisation: which covariates are really important to predict cumulative live birth and which models are best? Pract. Res. Clin. Obstetr. Gynaecol. 135 , 102309–102329 (2022).

Jenkins, J. et al. Empathetic application of machine learning may address appropriate utilization of ART. Reprod. BioMed. Online 41 , 573–577 (2020).

Senapati, S. et al. The fast track to fertility program: rapid cycle innovation to redesign fertility care. NEJM Catal. Innov. Care Deliv. 3 , CAT–22 (2022).

Google Scholar  

Mesko, B. & Topol, E. J. The imperative for regulatory oversight of large language models (or generative AI) in healthcare. NPJ Digi. Med. 6 , 120 (2023).

Article   Google Scholar  

Broekmans, F. J. Individualization of FSH doses in assisted reproduction: facts and fiction. Front. Endocrinol. 10 , 181 (2019).

Abbara, A. et al. FSH requirements for follicle growth during controlled ovarian stimulation. Front. Endocrinol. 10 , 579 (2019).

Abbara, A., Clarke, S. A. & Dhillo, W. S. Novel concepts for inducing final oocyte maturation in in vitro fertilization treatment. Endocr. Rev. 39 , 593–628 (2018).

Article   PubMed   PubMed Central   Google Scholar  

Fanton, M. et al. An interpretable machine learning model for individualized gonadotropin starting dose selection during ovarian stimulation. Reprod. BioMed. Online https://doi.org/10.1016/j.rbmo.2022.07.010 (2022).

Fanton, M., Baker, V. L. & Loewke, K. E. Selection of optimal gonadotropin dose using machine learning may be associated with improved outcomes and reduced utilization of FSH. Fertil. Steril. 118 , e80–e81 (2022).

Ferrand, T. et al. Predicting the number of oocytes retrieved from controlled ovarian hyperstimulation with machine learning. Hum. Reprod. 38 , 1918–1926 (2023).

Nguyen, T. et al. A novel decentralized federated learning approach to train on globally distributed, poor quality, and protected private medical data. Sci. Rep. 12 , 1–12 (2022).

Heinis, T. & Ailamaki, A. Data Infrastructure for Medical Research 2nd edn, Vol. 4 (Now Publishers, 2017).

Correa, N., Cerquides, J., Arcos, J. L. & Vassena, R. Supporting first FSH dosage for ovarian stimulation with machine learning. Reprod. BioMed. Online 45 , 1039–1045 (2022).

Xu, H. et al. POvaStim: An online tool for directing individualized FSH doses in ovarian stimulation. Innovation 4 , 100401 (2023).

CAS   PubMed   PubMed Central   Google Scholar  

Zieliński, K. et al. Personalized prediction of the secondary oocytes number after ovarian stimulation: A machine learning model based on clinical and genetic data. PLoS Comput. Biol. 19 , e1011020 (2023).

Abbara, A. et al. Endocrine requirements for oocyte maturation following hCG, GnRH agonist, and kisspeptin during IVF treatment. Front. Endocrinol. 764 , 412999 (2020).

Voliotis, M. et al. Quantitative approaches in clinical reproductive endocrinology. Curr. Opin. Endocr. Metabol. Res. 88 , 100421 (2022).

Wiens, J. et al. Do no harm: a roadmap for responsible machine learning for health care. Nat. Med. 25 , 1337–1340 (2019).

Andersen, A. N. et al. Individualized versus conventional ovarian stimulation for in vitro fertilization: a multicenter, randomized, controlled, assessor-blinded, phase 3 noninferiority trial. Fertil. Steril. 107 , 387–396 (2017).

Ishihara, O. & Arce, J.-C. et al. Individualized follitropin delta dosing reduces OHSS risk in Japanese IVF/ICSI patients: a randomized controlled trial. Reprod. Biomed. Online 42 , 909–918 (2021).

Qiao, J. et al. A randomised controlled trial to clinically validate follitropin delta in its individualised dosing regimen for ovarian stimulation in asian IVF/ICSI patients. Hum. Reprod. 36 , 2452–2462 (2021).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Abbara, A. et al. Follicle size on day of trigger most likely to yield a mature oocyte. Front. Endocrinol. 9 , 193 (2018).

Hariton, E. et al. A machine learning algorithm can optimize the day of trigger to improve in vitro fertilization outcomes. Fertil. Steril. 116 , 1227–1235 (2021).

Fanton, M. et al. An interpretable machine learning model for predicting the optimal day of trigger during ovarian stimulation. Fertil. Steril. 118 , 101–108 (2022).

Rudin, C. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nat. Mach. Intell. 1 , 206–215 (2019).

Afnan, M. A. M. et al. Interpretable, not black-box, artificial intelligence should be used for embryo selection. Hum. Reprod. Open (2021).

Letterie, G. & Mac Donald, A. Artificial intelligence in in vitro fertilization: a computer decision support system for day-to-day management of ovarian stimulation during in vitro fertilization. Fertil. Steril. 114 , 1026–1031 (2020).

Letterie, G., MacDonald, A. & Shi, Z. An artificial intelligence platform to optimize workflow during ovarian stimulation and IVF: process improvement and outcome-based predictions. Reprod. BioMed. Online 44 , 254–260 (2022).

Robertson, I., Chmiel, F. & Cheong, Y. Streamlining follicular monitoring during controlled ovarian stimulation: a data-driven approach to efficient IVF care in the new era of social distancing. Hum. Reprod. 36 , 99–106 (2021).

CAS   PubMed   Google Scholar  

Dimitriadis, I., Zaninovic, N., Badiola, A. C. & Bormann, C. L. Artificial intelligence in the embryology laboratory: a review. Reprod. BioMed. Online (2021).

Riegler, M. A. et al. Artificial intelligence in the fertility clinic: status, pitfalls and possibilities. Hum. Reprod. 36 , 2429–2442 (2021).

Finelli, R., Leisegang, K., Tumallapalli, S., Henkel, R. & Agarwal, A. The validity and reliability of computer-aided semen analyzers in performing semen analysis: a systematic review. Transl. Androl. Urol. 10 , 3069–3079 (2021).

Dearing, C., Jayasena, C. & Lindsay, K. Can the sperm class analyser (SCA) CASA-Mot system for human sperm motility analysis reduce imprecision and operator subjectivity and improve semen analysis? Hum. Fertil. (2019).

Shibahara, H. et al. Prediction of pregnancy by intrauterine insemination using CASA estimates and strict criteria in patients with male factor infertility. Int. J. Androl. 27 , 63–68 (2004).

Garrett, C., Liu, D., Clarke, G., Rushford, D. & Baker, H. Automated semen analysis: ‘zona pellucida preferred’ sperm morphometry and straight line velocity are related to pregnancy rate in subfertile couples. Hum. Reprod. 18 , 1643–1649 (2003).

Larsen, L. et al. Computer-assisted semen analysis parameters as predictors for fertility of men from the general population. Hum. Reprod. 15 , 1562–1567 (2000).

Organization, W. H. et al. WHO Laboratory Manual for the Examination and Processing of Human Semen 6th edn, Vol. 2 (World Health Organization, 2021).

Gallagher, M. T., Cupples, G., Ooi, E. H., Kirkman-Brown, J. C. & Smith, D. J. Rapid sperm capture: high-throughput flagellar waveform analysis. Hum. Reprod. 34 , 1173–1185 (2019).

Badura, A. et al. Prediction of semen quality using artificial neural network. J. Appl. Biomed. 17 , 167–174 (2019).

Vickram, A. S. et al. Validation of artificial neural network models for predicting biochemical markers associated with male infertility. Syst. Biol. Reprod. Med. 62 , 258–265 (2016).

Llavanera, M., Delgado-Bermúdez, A., Ribas-Maynou, J., Salas-Huetos, A. & Yeste, M. A systematic review identifying fertility biomarkers in semen: a clinical approach through omics to diagnose male infertility. Fertil. Steril. 118 , 291–313 (2022).

Hicks, S. A. et al. Machine learning-based analysis of sperm videos and participant data for male fertility prediction. Sci. Rep. 9 , 16770 (2019).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Thambawita, V., Halvorsen, P., Hammer, H., Riegler, M. & Haugen, T. B. Extracting temporal features into a spatial domain using autoencoders for sperm video analysis. arXiv (2019).

Ottl, S., Amiriparian, S., Gerczuk, M. & Schuller, B. W. motilitAI: A machine learning framework for automatic prediction of human sperm motility. iScience 25 , 104644 (2022).

Saiffe Farías, A. F. et al. Single-sperm motility analysis during ICSI using an artificial intelligence sperm identification software (SID) and correlation with morphology. Fertil. Steril. 118 , e56–e57 (2022).

Mendizabal-Ruiz, G. et al. Computer software (SID) assisted real-time single sperm selection associated with fertilization and blastocyst formation. Reprod. BioMed. Online 45 , 703–711 (2022).

You, J. B. et al. Machine learning for sperm selection. Nat. Rev. Urol. 18 , 387–403 (2021).

Chang, V., Garcia, A., Hitschfeld, N. & Härtel, S. Gold-standard for computer-assisted morphological sperm analysis. Comput. Biol. Med. 83 , 143–150 (2017).

Shaker, F., Monadjemi, S. A., Alirezaie, J. & Naghsh-Nilchi, A. R. A dictionary learning approach for human sperm heads classification. Comput. Biol. Med. 91 , 181–190 (2017).

Riordon, J., McCallum, C. & Sinton, D. Deep learning for the classification of human sperm. Comput. Biol. Med. 111 , 103342 (2019).

Zhang, Y. et al. Improving human sperm head morphology classification with unsupervised anatomical feature distillation. In 2022 IEEE 19th International Symposium on Biomedical Imaging (ISBI) 01–05 (IEEE, 2022).

Movahed, R. A., Mohammadi, E. & Orooji, M. Automatic segmentation of sperm’s parts in microscopic images of human semen smears using concatenated learning approaches. Comput. Biol. Med. 109 , 242–253 (2019).

Javadi, S. & Mirroshandel, S. A. A novel deep learning method for automatic assessment of human sperm images. Comput. Biol. Med. 109 , 182–194 (2019).

Abbasi, A., Miahi, E. & Mirroshandel, S. A. Effect of deep transfer and multi-task learning on sperm abnormality detection. Comput. Biol. Med. 128 , 104121 (2021).

Jiang, A., Jiaqi, W., Zhao, H., Zhang, Z. & Sun, Y. Identifying viability of immotile sperm at one glance: Sperm viability classifier powered by deep learning. Fertil. Steril. 118 , e297–e298 (2022).

Kresch, E., Efimenko, I., Gonzalez, D., Rizk, P. J. & Ramasamy, R. Novel methods to enhance surgical sperm retrieval: a systematic review. Arab J. Urol. 19 , 227–237 (2021).

Wu, D. J., Badamjav, O., Reddy, V. V., Eisenberg, M. & Behr, B. A preliminary study of sperm identification in microdissection testicular sperm extraction samples with deep convolutional neural networks. Asian J. Androl. 23 , 135–139 (2021).

Bachelot, G. et al. A machine learning approach for the prediction of testicular sperm extraction in nonobstructive azoospermia: algorithm development and validation study. J. Med. Inter. Res. 25 , e44047 (2023).

McCallum, C. et al. Deep learning-based selection of human sperm with high DNA integrity. Commun. Biol. 2 , 250 (2019).

Kuroda, S. et al. Development of a novel robust artificial intelligence developed sperm DNA fragmentation test—preliminary findings. Fertil. Steril. 118 , e307 (2022).

Peng, T. et al. Machine learning-based clustering to identify the combined effect of the DNA fragmentation index and conventional semen parameters on in vitro fertilization outcomes. Reprod. Biol. Endocrinol. 21 , 26 (2023).

Cissen, M. et al. Measuring sperm DNA fragmentation and clinical outcomes of medically assisted reproduction: a systematic review and meta-analysis. PLoS One 11 , e0165125 (2016).

Agarwal, A. et al. Sperm DNA fragmentation: a new guideline for clinicians. World J. Mens Health 38 , 412–471 (2020).

Esteves, S. C. et al. Sperm DNA fragmentation testing: summary evidence and clinical practice recommendations. Andrologia 53 , e13874 (2021).

Alahmar, A. T., Singh, R. & Palani, A. Sperm DNA fragmentation in reproductive medicine: a review. J. Hum. Reprod. Sci. 15 , 206–218 (2022).

Zaninovic, N. & Rosenwaks, Z. Artificial intelligence in human in vitro fertilization and embryology. Fertil. Steril. 114 , 914–920 (2020).

Rienzi, L. et al. Significance of metaphase ii human oocyte morphology on ICSI outcome. Fertil. Steril. 90 , 1692–1700 (2008).

Balaban, B. & Urman, B. Effect of oocyte morphology on embryo development and implantation. Reprod. BioMed. Online 12 , 608–615 (2006).

Kanakasabapathy, M., Bormann, C., Thirumalaraju, P., Banerjee, R. & Shafiee, H. P. Improving the performance of deep convolutional neural networks (CNN) in embryology using synthetic machine-generated images. In Human Reproduction 35th edn, Vol. 209 (Oxford University Press, 2020).

Kanakasabapathy, M. K. et al. Adaptive adversarial neural networks for the analysis of lossy and domain-shifted datasets of medical images. Nat. Biomed. Eng. 5 , 571–585 (2021).

Nayot, D., Meriano, J., Casper, R. & Alex, K. An oocyte assessment tool using machine learning; predicting blastocyst development based on a single image of an oocyte. Hum. Reprod. 35 , 129–130 (2020).

Mercuri, N., Fjeldstad, J., Krivoi, A., Meriano, J. & Nayot, D. A non-invasive, 2-dimensional (2D) image analysis artificial intelligence (AI) tool scores mature oocytes and correlates with the quality of subsequent blastocyst development. Fertil. Steril. 118 , e78–e79 (2022).

Link, C. et al. P-246 A novel non-invasive tool for oocyte selection using gene expression and artificial intelligence. Hum. Reprod. 37 , deac107–236 (2022).

Janati, S., Behmanesh, M. A., Najafzadehvarzi, H., Akhundzade, Z. & Poormoosavi, S. M. Follicular fluid zinc level and oocyte maturity and embryo quality in women with polycystic ovary syndrome. Int. J. Fertil. Steril. 15 , 197–201 (2021).

Cheng, E.-H. et al. Evaluation of telomere length in cumulus cells as a potential biomarker of oocyte and embryo quality. Hum. Reprod. 28 , 929–936 (2013).

Kirillova, A., Smitz, J. E. J., Sukhikh, G. T. & Mazunin, I. The role of mitochondria in oocyte maturation. Cells 10 , 2484 (2021).

Lemseffer, Y., Terret, M.-E., Campillo, C. & Labrune, E. Methods for assessing oocyte quality: a review of literature. Biomedicines 10 , 2184 (2022).

Dimitriadis, I. et al. Deep convolutional neural networks (CNN) for assessment and selection of normally fertilized human embryos. Fertil. Steril. 112 , e272 (2019).

Fukunaga, N. et al. Development of an automated two pronuclei detection system on time-lapse embryo images using deep learning techniques. Reprod. Med. Biol. 19 , 286–294 (2020).

Coticchio, G. et al. Cytoplasmic movements of the early human embryo: imaging and artificial intelligence to predict blastocyst development. Reprod. Biomed. Online 42 , 521–528 (2021).

Zhao, M. et al. Application of convolutional neural network on early human embryo segmentation during in vitro fertilization. J. Cell. Mol. Med. 25 , 2633–2644 (2021).

Khosravi, P. et al. Deep learning enables robust assessment and selection of human blastocysts after in vitro fertilization. NPJ Digi. Med. 2 , 1–9 (2019).

Thirumalaraju, P. et al. Evaluation of deep convolutional neural networks in classifying human embryo images based on their morphological quality. Heliyon 7 , e06298 (2021).

Berntsen, J., Rimestad, J., Lassen, J. T., Tran, D. & Kragh, M. F. Robust and generalizable embryo selection based on artificial intelligence and time-lapse image sequences. PLoS One 17 , e0262661 (2022).

Theilgaard Lassen, J., Fly Kragh, M., Rimestad, J., Nygård Johansen, M. & Berntsen, J. Development and validation of deep learning based embryo selection across multiple days of transfer. Sci. Rep. 13 , 4235 (2023).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Diakiw, S. M. et al. An artificial intelligence model correlated with morphological and genetic features of blastocyst quality improves ranking of viable embryos. Reprod. Biomed. Online 45 , 1105–1117 (2022).

Ahlström, A. et al. A double-blind randomized controlled trial investigating a time-lapse algorithm for selecting day 5 blastocysts for transfer. Hum. Reprod. 37 , 708–717 (2022).

Goodman, L. R., Goldberg, J., Falcone, T., Austin, C. & Desai, N. Does the addition of time-lapse morphokinetics in the selection of embryos for transfer improve pregnancy rates? a randomized controlled trial. Fertil. Steril. 105 , 275–285 (2016).

Kieslinger, D. C. et al. Clinical outcomes of uninterrupted embryo culture with or without time-lapse-based embryo selection versus interrupted standard culture (SelecTIMO): a three-armed, multicentre, double-blind, randomised controlled trial. Lancet 401 , 1438–1446 (2023).

Pribenszky, C., Nilselid, A.-M. & Montag, M. Time-lapse culture with morphokinetic embryo selection improves pregnancy and live birth chances and reduces early pregnancy loss: a meta-analysis. Reprod. Biomed. Online 35 , 511–520 (2017).

Hickman, C. et al. Turning the black box into a glass box: use of transparent artificial intelligence to understand biological markers useful for embryo selection. Fertil. Steril. 118 , e5–e6 (2022).

Hickman, C. et al. Comprehensive comparison of number of embryology hours per cycle and risk before and after introduction of CHLOE EQ™ (Fairtility) into a 100% time-lapse IVF clinic. Fertil. Steril. 118 , e119–e120 (2022).

Tiegs, A. W. et al. A multicenter, prospective, blinded, nonselection study evaluating the predictive value of an aneuploid diagnosis using a targeted next-generation sequencing-based preimplantation genetic testing for aneuploidy assay and impact of biopsy. Fertil. Steril. 115 , 627–637 (2021).

Wang, L. et al. IVF embryo choices and pregnancy outcomes. Prenat. Diagn. 41 , 1709–1717 (2021).

Hipp, H. S. et al. Trends and outcomes for preimplantation genetic testing in the United States, 2014-2018. JAMA 327 , 1288–1290 (2022).

Meseguer Escriva, M. et al. O-073 Artificial intelligence (AI) based triage for preimplantation genetic testing (PGT); an AI model that detects novel features in the embryo associated with ploidy. Hum. Reprod. 37 , deac104–087 (2022).

Barnes, J. et al. A non-invasive artificial intelligence approach for the prediction of human blastocyst ploidy: a retrospective model development and validation study. Lancet Digi. Health 5 , e28–e40 (2023).

Article   CAS   Google Scholar  

Chavez-Badiola, A., Flores-Saiffe-Farías, A., Mendizabal-Ruiz, G., Drakeley, A. J. & Cohen, J. Embryo ranking intelligent classification algorithm (erica): artificial intelligence clinical assistant predicting embryo ploidy and implantation. Reprod. BioMed. Online 41 , 585–593 (2020).

Jiang, V. S. et al. The use of voting ensembles and patient characteristics to improve the accuracy of deep neural networks as a non-invasive method to classify embryo ploidy status. Fertil. Steril. 116 , e155–e156 (2021).

Liang, R. et al. Prediction model for day 3 embryo implantation potential based on metabolites in spent embryo culture medium. BMC Pregn. Childbirth 23 , 425 (2023).

Eldarov, C. et al. LC-MS analysis revealed the significantly different metabolic profiles in spent culture media of human embryos with distinct morphology, karyotype and implantation outcomes. Int. J. Mol. Sci. 23 , 2706 (2022).

Vergouw, C. G. et al. No evidence that embryo selection by near-infrared spectroscopy in addition to morphology is able to improve live birth rates: results from an individual patient data meta-analysis. Hum. Reprod. 29 , 455–461 (2014).

Kirkegaard, K. et al. Nuclear magnetic resonance metabolomic profiling of day 3 and 5 embryo culture medium does not predict pregnancy outcome in good prognosis patients: a prospective cohort study on single transferred embryos. Hum. Reprod. 29 , 2413–2420 (2014).

Lledo, B., Morales, R., Antonio Ortiz, J., Bernabeu, A. & Bernabeu, R. Noninvasive preimplantation genetic testing using the embryo spent culture medium: an update. Curr. Opin. Obstet. Gynecol. 35 , 294–299 (2023).

Siristatidis, C. S., Sertedaki, E., Vaidakis, D., Varounis, C. & Trivella, M. Metabolomics for improving pregnancy outcomes in women undergoing assisted reproductive technologies. Cochr. Datab. Syst. Rev. 3 , CD011872 (2018).

Cheredath, A. et al. Combining machine learning with metabolomic and embryologic data improves embryo implantation prediction. Reprod. Sci. 30 , 984–994 (2023).

Siristatidis, C. et al. Why has metabolomics so far not managed to efficiently contribute to the improvement of assisted reproduction outcomes? the answer through a review of the best available current evidence. Diagnost. Basel 11 , 1602 (2021).

Doyle, N. et al. Live birth after transfer of a single euploid vitrified-warmed blastocyst according to standard timing vs. timing as recommended by endometrial receptivity analysis. Fertil. Steril. 118 , 314–321 (2022).

Richter, K. S. & Richter, M. L. Personalized embryo transfer reduces success rates because endometrial receptivity analysis fails to accurately identify the window of implantation. Hum. Reprod. 38 , 1239–1244 (2023).

(The writing group) for the participants to the 2022 Lugano RIF Workshop. Recurrent implantation failure: reality or a statistical mirage? Consensus statement from the July 1, 2022 Lugano workshop on recurrent implantation failure. Fertil. Steril. 120 , 45–59 (2023).

Gromski, P. S. et al. Ethnic discordance in serum anti-müllerian hormone in european and indian healthy women and indian infertile women. Reprod. Biomed. Online 45 , 979–986 (2022).

Ko, J. K. et al. Comparison of the number of oocytes obtained after ovarian stimulation between chinese and caucasian women undergoing in vitro fertilization using a standardized stimulation regime. J. Ovarian Res. 14 , 175 (2021).

Loutradis, D. et al. FSH receptor gene polymorphisms have a role for different ovarian response to stimulation in patients entering IVF/ICSI-ET programs. J. Assist. Reprod. Genet. 23 , 177–184 (2006).

Roth, L. W. et al. Evidence of GnRH antagonist escape in obese women. J. Clin. Endocrinol. Metab. 99 , E871–E875 (2014).

Venetis, C. A. et al. What is the optimal GnRH antagonist protocol for ovarian stimulation during ART treatment? A systematic review and network meta-analysis. Hum. Reprod. Update (2023).

Garg, A. et al. Luteal phase support in assisted reproductive technology. Nat. Rev. Endocrinol. (2023).

Amann, J. et al. To explain or not to explain?-artificial intelligence explainability in clinical decision support systems. PLoS Digi. Health 1 , e0000016 (2022).

Vasey, B. et al. Reporting guideline for the early-stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI. Nat. Med. 28 , 924–933 (2022).

Collins, G. S. et al. Protocol for development of a reporting guideline (TRIPOD-AI) and risk of bias tool (PROBAST-AI) for diagnostic and prognostic prediction model studies based on artificial intelligence. BMJ Open 11 , e048008 (2021).

Curchoe, C. L. Proceedings of the first world conference on AI in fertility. J. Assist. Reprod. Genet. 40 , 215–222 (2023).

Joshi, K. et al. A proof-of-concept prospective study of applying artificial intelligence for sperm selection in the IVF laboratory. Reprod. Reprod. BioMed. Online 188 , 103329 (2023).

Download references

Acknowledgements

The Department of Metabolism, Digestion, and Reproduction is funded by grants from the MRC and NIHR. S.H. is supported by the UKRI CDT in AI for Healthcare http://ai4health.io (EP/S023283/1). A.A. is supported by an NIHR Clinician Scientist Award (CS-2018-18-ST2-002). M.V. and K.T.A. are supported by the EPSRC (EP/T017856/1). W.S.D. is supported by an NIHR Senior Investigator Award (NIHR202371).

Author information

These authors contributed equally: Simon Hanassab, Ali Abbara.

These authors jointly supervised this work: Thomas Heinis, Waljit S. Dhillo.

Authors and Affiliations

Department of Metabolism, Digestion, and Reproduction, Imperial College London, London, UK

Simon Hanassab, Ali Abbara, Arthur C. Yeung, Geoffrey H. Trew & Waljit S. Dhillo

Department of Computing, Imperial College London, London, UK

Simon Hanassab & Thomas Heinis

UKRI Centre for Doctoral Training in AI for Healthcare, Imperial College London, London, UK

Simon Hanassab

Imperial College Healthcare NHS Trust, London, UK

Ali Abbara, Arthur C. Yeung & Waljit S. Dhillo

Department of Mathematics and Statistics, University of Exeter, Exeter, UK

Margaritis Voliotis & Krasimira Tsaneva-Atanasova

Living Systems Institute, University of Exeter, Exeter, UK

EPSRC Hub for Quantitative Modelling in Healthcare, University of Exeter, Exeter, UK

School of Computer Science, University of St Andrews, St Andrews, UK

Tom W. Kelsey

The Fertility Partnership, Oxford, UK

Geoffrey H. Trew & Scott M. Nelson

School of Medicine, University of Glasgow, Glasgow, UK

Scott M. Nelson

Biomedical Research Centre, University of Bristol, Bristol, UK

You can also search for this author in PubMed   Google Scholar

Contributions

S.H., A.A., T.H., and W.S.D. conceptualized the review. S.H., A.A., A.C.Y., M.V., and S.M.N. wrote the manuscript. M.V., K.T.A., T.W.K., and T.H. provided methodological expertise. A.A., A.C.Y., G.H.T., S.M.N., and W.S.D. provided clinical expertise. All authors reviewed and approved the final manuscript.

Corresponding author

Correspondence to Waljit S. Dhillo .

Ethics declarations

Competing interests.

A.A. has received grants from the BRC; and has provided consulting services for Myovant Sciences Ltd. G.H.T. has stock in TFP; has received honoraria and travel support from Ferring Pharmaceuticals; and has provided consultancy services to ARC Medical Inc. S.M.N. received grants from NIHR, CSO, and BRC; provided consultancy services for Access Fertility, Modern Fertility, TFP, and Ferring Pharmaceuticals; received honoraria from Ferring Pharmaceuticals and Merck; received support for attending meetings and/or travel from Ferring Pharmaceuticals and Merck; participated in a data safety monitoring board or advisory board for NIHR; owns stock or stock options in TFP. W.S.D. received grants from NIHR, MRC, and Imperial Health Charity, and is a Consultant for Myovant Sciences Ltd. The remaining authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Hanassab, S., Abbara, A., Yeung, A.C. et al. The prospect of artificial intelligence to personalize assisted reproductive technology. npj Digit. Med. 7 , 55 (2024). https://doi.org/10.1038/s41746-024-01006-x

Download citation

Received : 25 January 2023

Accepted : 10 January 2024

Published : 01 March 2024

DOI : https://doi.org/10.1038/s41746-024-01006-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Predictive strategies for oocyte maturation in ivf cycles: from single indicators to integrated models.

Middle East Fertility Society Journal (2024)

A review of artificial intelligence applications in in vitro fertilization

• Xiaowen Liang

Journal of Assisted Reproduction and Genetics (2024)

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

Main topics in assisted reproductive market: A scoping review

Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing – original draft

Affiliations Applied Molecular Biology Lab (LAPLIC), Biochemistry Department, Federal University of Rio Grande do Norte (UFRN), Natal, Rio Grande do Norte, Brazil, Januário Cicco´s University Hospital (MEJC), Federal University of Rio Grande do Norte (UFRN), Natal, Rio Grande do Norte, Brazil

Roles Data curation

Affiliation Applied Molecular Biology Lab (LAPLIC), Biochemistry Department, Federal University of Rio Grande do Norte (UFRN), Natal, Rio Grande do Norte, Brazil

Roles Conceptualization, Formal analysis, Funding acquisition, Investigation, Project administration, Supervision, Visualization, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

• Janaina Ferreira Aderaldo, 
• Beatriz Helena Dantas Rodrigues de Albuquerque, 
• Maryana Thalyta Ferreira Câmara de Oliveira, 
• Mychelle de Medeiros Garcia Torres, 
• Daniel Carlos Ferreira Lanza

• Published: August 1, 2023
• https://doi.org/10.1371/journal.pone.0284099
• Reader Comments

Infertility affects around 12% of couples, and this proportion has been gradually increasing. In this context, the global assisted reproductive technologies (ART) market shows significant expansion, hovering around USD 26 billion in 2019 and is expected to reach USD 45 billion by 2025.

We realized a scoping review of the ART market from academic publications, market reports, and specialized media news, to identify the main terms and characterize them into the main topics in the area.

We apply an LDA topic modeling process to identify the main terms, and clustered them into semantic synonymous topics. We extracted the patterns and information to these topics and purposed a factor/consequence correlation to them.

We found 2,232 academic papers and selected 632 to include in the automatic term detection. We also included 34 market reports and seven notices produced by specialized enterprises. Were identified 121 most relevant cited terms covering 7,806 citations. These terms were manually aggregated into 10 topics based on semantic similarity: neutral terms (37.2%), economic aspects (17.6%), in vitro fertilization ( IVF) commodities & cross-border reproductive care (CBRC) (10.6%), geographic distribution (9.5%), social aspects (7%), regulation (6%), trends & concerns (3.9%), accessibility (3.4%), internet influence (2.9%), and fertility preservation for non-medical reasons (2%).

The analysis indicates a market with expressive complexity. Most terms were associated with more than one topic, indicating the synergism of this market’s behavior. Only seven terms related to economic aspects, surrogacy and donation represent around 50% of the citations. Except for the topic formed by generic terms, the topic of the economic aspects was the most represented, reflecting macro perspectives such as a-la-carte standard of treatments, many clinics operating on a small/medium scale, and the recent formation of conglomerates. The IVF commodities & CBRC topic brings an overview of gametes pricing and transnational surrogacy, and its regulation. The topic of geographic distribution indicates that that the Asia-Pacific (APAC) market has the most significant growth potential in all fields. Despite the increase in supply and demand for infertility treatments and technological advances in recent decades, the success rate of IVF cycles remains at around 30%. Terms referring to research and development or technical improvement were not identified in a significant way in this review.

Conclusions

The formation of topics by semantic similarity proved to be an initial path for the elaboration of in-depth studies on the dynamics between several factors, for this, we present the panel classifying main terms into factors (demand, pent-up demand, or distributive) or ART market consequences. Through this approach, it was possible to observe that most of the works addresses economic aspects, regulation and geographic aspects and that topics related to research and improvement have not been addressed. In this way, we highlight the need to deepen the analysis of market elements that may be related to increased efficiency of IVF in the technical field.

Citation: Aderaldo JF, Rodrigues de Albuquerque BHD, Câmara de Oliveira MTF, de Medeiros Garcia Torres M, Lanza DCF (2023) Main topics in assisted reproductive market: A scoping review. PLoS ONE 18(8): e0284099. https://doi.org/10.1371/journal.pone.0284099

Editor: Meijia Zhang, China Agricultural University, CHINA

Received: June 28, 2022; Accepted: March 19, 2023; Published: August 1, 2023

Copyright: © 2023 Aderaldo et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: This study was funded by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript

Competing interests: The authors declare that they have no competing interests.

Introduction

Infertility is currently defined as a failure to achieve a clinical pregnancy after at least one year of regular attempts [ 1 ], which was later updated to include physiological or psychological conditions incompatible with natural meeting of gametes [ 2 ]. It affects between 8% and 12% of couples globally [ 3 ], and this proportion is gradually increasing [ 1 , 4 ] because of multiple causes, such as the modern lifestyle, diseases, and the postponement of parenthood [ 5 – 7 ].

In that context, it is estimated that half of the infertile couples never seek fertility [ 8 ], and the investigation of the reasons reveals a complex product of the national public and private health policies and economic, political, and cultural factors [ 9 – 11 ].

However, the global assisted reproductive technology (ART) market expanded in clinic numbers and procedures [ 9 ], The ART market services were around USD 26 billion in 2019 [ 12 ] and are expected to reach USD 45 billion by 2025 [ 5 ]. Between 1997 and 2016, ART treatments have increased more than five-fold in Europe, 4.6-fold in North America, and three-fold in Australia and New Zealand [ 13 , 14 ].

The distribution of this billionaire market is heterogeneous due to complex clustering factors like unequal regulatory restrictions, local procedures practices, and socio-cultural differences associated with disposable income [ 7 , 15 – 18 ]. Despite the volume of information about several factors that compose this market, there is no structured analysis of these factors in clusters.

For this reason, we produce this scoping review for the identification of the main terms and topics cited in ART market texts. This is an appropriate tool for examining emerging evidence that has not been comprehensively reviewed or of a complex and heterogeneous nature, mapping the available evidence for clarifying definitions and conceptual boundaries [ 19 , 20 ].

The following question guided this review: What are the aspects that compose the global ART market?

To answer the question that guided this review, we choose the scoping review approach with Latent Dirichlet Allocation (LDA) topic modeling as the method to identify this evidence.

We performed a scoping review based on guidelines proposed by the Joanna Briggs Institute (JBI) Scoping Review Methodology Group [ 21 ]. The methodology was adapted from Tricco et al . (2017) [ 22 ]: a) elaboration of the research question; b) identification of relevant studies; c) selection of relevant terms by LDA topic modeling using an automatic tool and aggregation of them by iterative team approach for studying a selection and data extraction [ 23 , 24 ]; d) chart production from the data incorporating quantitative and qualitative thematic analysis; e) summarization and report of the results ( Fig 1 ).

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

This design was adapted from Page et al. (2020) and purposes to identify the studies in databases and other methods through the query elaborated from the research question. After selecting studies by eligibility criteria, the titles, keywords, and abstracts were subjected to detecting terms using the topic modeling approach [ 23 ]. Terms were aggregated by semantic similarity into clusters (topics), summarized, and presented.

https://doi.org/10.1371/journal.pone.0284099.g001

Eligibility criteria

We included the following peer-reviewed and gray literature in this review: a) academic publications; b) market reports about the ART market produced by specialized research companies; and c) selected referenced media news.

About peer-reviewed publications, the following bibliographic databases were screened from 2010 to 2022: PUBMED, MEDLINE, EMBASE and Google Scholar. We defined the query (((assisted reproductive market) OR (infertility market)) OR (fertility market)) AND (("01/01/2010"[Date—Publication]: "2022"[Date—Publication]))).

For gray literature, market reports, and media news, we searched Google using the terms ‘assisted reproductive market’, which focused on referenced economic agency websites, and specialized media websites, excluding blogs and clinic websites.

Search and selection of sources of evidence

Two independent reviewers selected pertinent literature through abstracts and titles using the Sysrev software [ 25 ]. Disagreements were resolved by consensus. After this selection, the selected studies were submitted to LDA topic modeling from the content of abstracts, title, and keywords using keywords Knime software [ 24 ], which identifies repetitive word patterns across a corpus of documents [ 26 ].

Synthesis of results

The terms detected by topic modeling were clustered by synonymous and semantic similarity into topic groups. We evaluate the content of these topics and present them in a quantitative approach by ranking the number of term citations in each topic about the recent ART, and a qualitative approach through an analytic mini review.

Methodological quality appraisal

We did not appraise methodological quality or risk of bias of the included articles, consistent with guidance on scoping review conduct [ 20 ]. We draw the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) checklist [ 27 ] in S1 Table .

Selection of sources of evidence

We found 2,232 academic papers and selected 632 that were eligible. We also included 34 market reports and seven notices produced by specialized enterprises. In these 673 records, were identified 121 most relevant cited terms covering 7,806 citations ( Fig 2 ). The academic evidence source represents 93.9% of the total, whose proportion remained approximate in the abundance of citations for each term.

We identified 2,232 academic publications in initial search. After duplicates, and ineligible exclusions (n = 1,600), we included 41 gray literature records. In total, were selected 673 abstracts, titles, and keywords for topic modeling screening step.

https://doi.org/10.1371/journal.pone.0284099.g002

The 121 identified terms were cited 7,806 times in 673 texts used for the terms detection approach ( Fig 3A and S2 Table ). The ratios were 0.18 for terms/number of texts and 11.6 for the number of citations/number of texts. We manually aggregated by team consensus these 121 terms into 10 clusters by semantic similarity ( Fig 3B and Table 1 ). As an example, the terms `ethical`and `social`were clustered into the topic of the social aspects.

A) 30 most cited terms detected by the topic modeling automation tool and each number of citations on the database. 26.9% of citations correspond to the neutral term reproductive . This was overestimated for being present in all titles and keywords and repeated in abstracts. Disregarding this term, we identified specific terms such as industry (586 citations) and surrogacy (345 citations); B) 10 Clusters of 121 detected terms aggregated by semantic similarity. We chose to organize the clusters based on the total number of citations (orange bars). The number of topics in each cluster is available in the blue bars.

https://doi.org/10.1371/journal.pone.0284099.g003

https://doi.org/10.1371/journal.pone.0284099.t001

The neutral terms topic corresponds to 37.2% of citations (27 terms; 2,900 citations). Despite this topic presenting the majority cluster, these terms were disregarded for the content analysis because they returned practically the entire set of records used.

The economic aspects topic corresponds to 17.6% of citations (16 terms; 1,375 citations) and refers to the econometric analysis of the market. We identified two subgroups, which following: a) a group of nine generic terms (1,178 citations) that returned unspecified records, and b) a group of seven terms (197 citations) representing specific economic terms such as insurance and coverage.

The compensation for reproductive services topic corresponds to 10.6% of citations (9 terms; 829 citations) and refers to records about the pricing of reproductive services and their ramifications. We identified two subgroups: a) commercial surrogacy (4 terms; 574 citations), and b) gametes pricing (5 terms; 255 citations).

The geographic distribution topic corresponds to 9.5% of citations (12 terms; 739 citations) and refers to the global distribution of the market in terms of size, demand, characteristics, and profitability.

The social aspects topic corresponds to 7% of citations (14 terms; 546 citations). In this group, the terms can be clustered into five subgroups, which follows: a) ethical/moral discussions (4 terms, 314 citations); b) religious issues (3 terms, 84 citations); c) gender issues (3 terms, 82 citations); d) sexual preferences (2 terms, 36 citations); and e) stress that was considered an independent term (10 citations).

The regulation topic corresponds to 6% of citations (9 terms; 469 citations) and refers to legal aspects of the ART market. It includes the assessment of the impact of regulation (laws and guidelines) on market behavior and the assessment of the impact of transnational practices on national regulation.

The ART market projections topic corresponds to 3.9% of citations (13 terms; 304 citations) and refers to prognostic analyzes. The group can be divided into three subgroups: positive trends (5 terms; 115 citations), b) concerns (5 terms; 152 citations), and c) neutral comparisons (3 terms; 47 citations).

The accessibility topic corresponds to 3.4% of citations (9 terms; 264 citations) and refers to terms that returned a set of texts to evaluate the social impacts related to the ART market.

The internet influence topic corresponds to 2.9% of citations (7 terms; 224 citations) and refers to terms that returned a set of texts related to the evaluation of the internet in the ART market.

The fertility preservation for non-medical reasons topic corresponds to 2% of citations (5 terms; 156 citations) and refers to terms that returned a set of texts related to this specific theme.

Critical appraisal within sources of evidence

Although resulting from the same search terms, we have found two essential differences in the content of academic publications and market studies. Academic publications focus, in general, on only one factor as a central objective and make an in-depth analysis. In contrast, market reports focus on economic projections and, more often, present the factors in a superficial and aggregated way as an increase/decrease ART market factor.

The specialized media news provided relevant information on the formation of conglomerates and open market companies. Considering that 93.9% of used literature was composed of peer-reviewed publications, the gray literature can provide a complementary perspective to peer-review information [ 28 ].

Summary of evidence

The ratio of 0.18 between the number of terms/number of texts and 11.6 between the number of citations/number of texts inform us that the texts generally deal with specific themes, whose citations are reinforced throughout the texts. Of the 121 terms detected, 16 were identified as variations of the same lexical root, and the others were identified as synonyms. We checked the accuracy of the automated tool by comparing the articles that gave rise to the identification of each of the terms. For many of these terms, there was practically total overlap between the groups of articles that gave rise to the terms identified as synonyms.

Neutral topics.

The terms grouped in neutral topics were disregarded from the content analysis because they practically returned all the records used in the term detection approach. This result is consistent with what was expected for the technique that uses the formation of matrices to search for terms and, for this reason, unspecific terms have a larger data set.

Economic aspects.

This topic refers mainly that the macro aspects of the market such as the mostly a la carte standard of treatments [ 29 ] and a structure with many clinics operating on a small/medium scale of the market that is rivaling the recent and growing formation of conglomerates [ 30 ].

Economic aspects are hardly quantifiable in imperfect markets such as health, where there are high interference levels from variables and market regulations [ 10 ]. The high prices are only one of several factors determining the ART market, which has cultural and religious associations that cannot be easily measured or evaluated by econometry [ 31 ]. However, the data collected on the growth of the ART market size in the last decades indicates that the regional discrepancies are derived from the different attractiveness for the several capital contributions made by different public and private subjects, a phenomenon known as the ’Matthew effect’ [ 32 ].

About the stock exchange and merging & acquisitions in the ART market (MAART), a few large companies have spent millions of dollars consolidating a fragmented IVF market [ 33 , 34 ]. While the conglomerates are growing, more venture capital firms invest in startups and fertility clinics, including specific niches [ 30 ]. These expansions reach state and national borders with a more entrepreneurial and corporate bias and heavy investments in technology [ 35 ].

Among the main actions carried out by companies in the sector we can mention:

• 2013—An Australian IVF company became the first IVF company traded on a major stock exchange, and it holds about 35% of the market [ 36 , 37 ].
• 2016—Cooper Surgical acquired Wallace Pharmaceuticals (India) for approximately USD 168 million [ 38 ].
• 2017—PitchBook accounted for more than US$ 178 million invested in startups that develop fertility products [ 34 ].
• 2017—The merger of IVI-RMA made this company the largest assisted reproduction center worldwide [ 39 ].
• 2017—The Thomson Medical Group Ltd. (TMG) formalized a joint venture to expand the IVI-RMA network in APAC and Mexico markets [ 40 ].
• 2019 –An enterprise that manages fertility benefits for employees of large companies reached USD 103.4 million in the first semester and released the shares on NASDAQ [ 41 ].

Coverage has a significant effect on use for older and more educated women, more significant than the effects found for other groups [ 42 , 43 ]. Studies report that more than half of working women consider changing jobs for better reproductive health benefits [ 44 ].

On the other hand, there is the possibility that insurance coverage laws may have adverse effects on total fertility in the medium and long term due to overly optimistic perceptions about the possibility of extending or delaying reproductive life in an induced way, which can be called ex-ante moral hazard [ 42 , 45 ], one of the alleged reasons for reducing public funding in Germany and Australia [ 46 , 47 ].

There is a growth in coverage for infertility treatment among jumbo employers, who tend to be trendsetters for smaller employers, and studies reported that more than half of working women would consider changing jobs for better reproductive health benefits [ 48 , 49 ]. On the other hand, there is the possibility that policies may have adverse effects on total fertility due to overly optimistic perceptions about the delay of reproductive life [ 42 , 45 ], one of the alleged reasons for reducing public funding in Germany and Australia [ 46 , 47 ].

Generally, economic recessions impact natural fertility in the developed world in does not leave a visible mark on the fertility levels of the global cohort [ 50 ]. The expressive increase in COVID-19 cases and massive hospitalizations has collapsed most health systems globally and caused the suspension of new fertility treatments, except for patients on cycle or who urgently require fertility preservation for oncological reasons [ 51 ].

Although the countries reacted with diverse responses in this pandemic, the ART services have been mainly responsive to public health and individual patient concerns [ 52 ]. The pandemic impact on fertility appears to have five main factors: high mortality, restricted access to family planning services, reduced work-life balance, economic recession and uncertainty, and disruptions to assisted reproduction services [ 53 ]. It is still early to assess how the pandemic caused by the Covid-19 disease has affected the ART market; however, it is expected that the economic recession and uncertainty impact assisted reproduction services.

Compensation for reproductive services.

Regarding the topic of c ompensation for reproductive services , the separation of these topics, although practical, has limitations because all of them are also strongly related to social aspects and legislation. We found two main analyses in the returned records for this topic: a) gamete pricing, and b) commercial surrogacy. Both are part of a more focused analysis on transnational markets called cross-border reproductive care (CBRC), popularly called reproductive tourism.

About 10% of IVF cycles are performed in the USA with donor eggs [ 54 ], and the results are like the use of fresh and frozen oocytes [ 55 ]. The term "donation" of gametes is considered inappropriate because they are generally sold [ 56 ]. The United Kingdom limits gamete’s values, while gamete donations are banned in Japan [ 54 ]. A complex set of stereotypes has led to the monetization of gametes and embryos and rapid response to price stratification based on donor phenotype and social characteristics as a degree or artistic achievements [ 57 – 59 ].

The CBRC is a global billionaire industry phenomenon that involves the transnational laissez-faire regulation [ 60 – 62 ], inequalities [ 63 ], and the demand for reproductive services [ 38 , 64 ]. It is a contentious and largely unregulated area [ 65 ] governed by the heterogeneity of conditions in each country [ 66 – 68 ]. At least ten motivations for CBRC have already been identified, grouped into four broad categories: legal and religious prohibitions, resource considerations; quality and safety concerns; and personal preferences [ 69 ].

Geographic distribution.

The topic of g eographic distribution comprises 9.5% of the citations (12 terms, 739 citations), the most discussed subject in the market reports. In general, the data presented addresses:

• The size of the market in billions of dollars: globally was around USD 26 billion in 2019 [ 12 ] and is expected to reach USD 45 billion by 2025 [ 5 ];
• Percentage distribution of clinics and number of procedures worldwide: Europe and North America represent ∼65% of the global ART market, followed by APAC with ∼25%; Middle East, Africa, and Latin America (also called by ’rest of the world’—RoW) representing ∼10% [ 5 ];
• Procedures and clinics per region: between 1997 and 2016, ART treatments have increased more than five-fold in Europe, 4.6-fold in North America, three-fold in Australia and New Zealand [ 13 , 14 ], with grown expectative in all scenarios and
• Factors (social/legal/economic) that impact this distribution: increasing infertility rates [ 7 , 15 , 16 , 38 , 70 , 71 ], rising disposable incomes [ 5 , 70 – 73 ], adoption of the western lifestyle [ 16 , 73 – 75 ], late family planning [ 16 , 70 , 72 ], low-cost and high-quality healthcare [ 18 , 72 , 75 ], favorable government initiatives [ 7 , 38 , 76 ], expansion of healthcare infrastructure [ 64 , 74 ], reduced socio-ethical stigma [ 15 , 77 ], and the CBRC [ 18 , 38 , 71 , 72 , 78 , 79 ].

ART market projections.

After neutral terms, the topic of ART market projections focused on more generic terms ( Table 1 ). The content mainly presents forecasts of the contents present in other topics such as social aspects, geographic distribution, accessibility, and regulation. The clustered terms comprised various database content, with analytical content as a characteristic in common. In addition to the market’s financial growth expectations, there is also an assessment of the geographic distribution, with the unanimous affirmation that the Asia-Pacific (APAC) market has the greatest compound annual growth rate (CAGR) and potential [ 5 , 7 , 15 , 16 , 38 , 70 – 75 , 77 , 78 ] ( S3 Table ).

The stock exchanges participation and mergers & acquisitions in the ART market (MAART) are a trend observed for a few large companies that have spent millions of dollars to consolidate a fragmented IVF market [ 33 , 34 ], with heavy investments in technology [ 35 ]. These companies also have been focused on specific niches considered non-traditional families [ 30 ].

The main projected concerns relate to reproductive commodification, in particular commercial surrogacy, and stereotypic gamete pricing [ 63 , 68 , 80 – 82 ]. In the same way that India regulated the issue to protect vulnerable women groups [ 65 , 83 ], there is a debate about ways of fair compensation for domestic surrogacy in Australia, the introduction of professional intermediaries, and limits on advertising to minimize risks [ 84 ]. It is an issue that is difficult to resolve and that depends on efforts and intranational agreements.

Social aspects.

The topic s ocial aspects subgroups can be clustered into five subgroups: a) ethical/moral discussions (4 terms, 314 citations); b) religious issues (3 terms, 84 citations); c) gender issues (3 terms, 82 citations); d) and sexual preferences (2 terms, 36 citations); and e) stress was considered an independent term (10 citations).

The records returned in this topic showed considerable overlap with the records returned in the accessibility topic, which is understandable because inequities are strongly associated with social and cultural characteristics [ 85 , 86 ].

It is complex to measure these social aspects’ impact on the ART market, a complex and imperfect health business where there are high interference levels [ 10 ]. The high prices are only one of several factors determining the ART market, which has cultural and religious associations that cannot be easily measured or evaluated by econometry [ 31 ].

It is estimated that a 1% increase in European national gross domestic product (GDP) would be able to increase 382 ART cycles per million women of reproductive age and, even so, it only increases 25% of this potential, concluding it is due to the social factors involved [ 87 ]. These factors also affect nations’ repayment policy (pro-natal or anti-natal) regardless of their GDPs, reflecting cultural and social priorities [ 88 ].

Many records about queer reproductive justice (QRJ) are returned on this topic. It refers to non-normative audiences who want to form a family nucleus, such as homo-affective couples, single parents, and other audiences who are discouraged when seeking reproductive services [ 89 , 90 ]. This market niche is often not directly related to accessibility and cost problems, and its acceptance has been partly driven globally by the strength of the neoliberal market [ 91 , 92 ].

Regulation.

Despite representing 6.7% of the number of citations in the detection stage, the topic of regulation represented 13% of the corresponding bibliography. Around 85 of those contributing to the IFFS triennial publication have regulated legislation or guidelines [ 65 ], generally associated with national economic and moral concerns [ 93 ].

We compared the data presented on the main modification to legislation in the last three years with the main topics presented by IFFS [ 65 ] detected, to assess whether there is synchronism in the topics assessed in legislation and academics ( Fig 4 ). We found some common points, but no direct correlation in the timing of the discussions.

The main themes of recent legislative changes stated by the IFFS (2019) coincide with the terms identified by topic modeling. In this way, we identified in which topics the similar terms were inserted and prepared this graph to illustrate the location of the main themes of legislative changes in relation to the main topics in assisted reproduction. Note that each topic may be included in more than one topic due to the complexity of the market.

https://doi.org/10.1371/journal.pone.0284099.g004

There is a recent debate on global policy and systematic regulatory forecasts to guide government responses to the existing market, preferably including a discussion open to all interested parties [ 94 , 95 ]. Most of the articles cited, regardless of the central focus, conclude with the statement of the need for consensual regulation at a global level to regulate the market and the public, thus avoiding most ethical conflicts [ 61 , 63 , 68 , 96 , 97 ].

Based on this lack of regulation, incoherent or fragmented regulation, the ART market worldwide is provided by free-market initiatives [ 63 ] and is associated with themes such as CBRC [ 98 ] and embryo gender selection [ 79 , 99 ]. Although it is a consensus that current regulations do not guarantee the exercise of reproductive rights and equal opportunities [ 100 , 101 ].

Accessibility.

The topic of accessibility has the same number of terms as the regulation topic but almost half the number of citations (9 terms, 264 citations). It is a topic strongly associated with economic aspects and regulation, as noted in the definition itself, which is the level of access to medical treatments necessary for infertility care through vantage/disadvantage in other aspects such as financial/social, race, class, gender, culture, and legal status played out on a social field [ 67 , 86 ].

Two terms that compile the dynamic and synergistic balance in the ART market, both domestic and transnational, are reproductive governance and stratified reproduction. Reproductive governance defines how different social actors use their powers to produce reproductive behaviors, such as legislative controls or permissions, economic incentives or disincentives, ethical and moral injunctions, or inductions [ 102 ]. Stratified reproduction refers to the inequity in reproductive rights by race, class, gender, culture, and legal status played out on a social field [ 67 ] that generates the accessibility and treatment offered to people into separate groups [ 85 , 86 ].

Accessibility, the ratio of the cost of IVF treatment to annual income [ 88 ], affects not only who can have access to ART treatment but also a) which treatments are used, as cheaper techniques are generally more likely to be covered by health insurance 100, and b) how ART is practiced, such as the association between accessibility and the number of embryos transferred [ 46 ]. This cascade of decisions impacts the results [ 10 ], and, still, most patients bear partial costs [ 103 ].

In IFFS 2019 Vigilance, 62% of the countries reported no existing family concept ART requirements; however, 50% reported limiting access to diagnostic or treatments mainly to single women or same-sex couples, excluding single men and intersex or transgender subjects [ 65 ]. This market niche is often not directly related to cost, and its acceptance has been partly driven globally by the strength of the neoliberal market [ 91 , 92 ].

Internet influence.

The topics of Internet influence and fertility preservation for non-medical reasons or social egg freezing (SEF) comprise less than 3% of the citations each. However, both are frequently cited in the texts included in the topic trends & concerns , where both growing trends and sources of concern and attention are pointed out. It is common sense that the internet and social media are powerful tools of massive influence, used by most patients during their infertility journey [ 104 , 105 ]. The content of these sites influences consumers’ selection process of both the chosen clinic and the doubts and desires for treatment and the possibility of high expectations [ 106 ]. The Society for Assisted Reproductive Technology (SART) updated your policy in 2018 to reduce public misunderstandings caused by different interpretations of data provided by clinics [ 105 ].

In addition to the search for information, the internet and social media have become spaces for selling surrogacy services in countries with legal permission or omission. This happens through forums for possible substitutes and customers [ 63 ] and a rapidly growing market for SEF and human commodities. While the benefit of dissemination and information is clear, it is essential to ensure that there is no misrepresentation and distribution of misleading information [ 107 ].

Fertility preservation for non-medical reasons.

Despite being the least represented among the topics (5 terms, 156 citations), s ocial freezing is one of the main trends [ 108 ]. Some jumbo enterprises announced the social freezing as a workplace benefit [ 48 , 49 ], although the American Society for Reproductive Medicine (ASRM) guideline includes a caution to avoid false hopes about delaying procreation [ 104 ]. The main reasons are not having a committed [ 109 , 110 ] , searching for financial security via career, or completing studies [ 111 , 112 ].

The possibility of preserving fertility in healthy women as a precaution for future infertility has gained strength in recent years [ 108 ] and the case of reproductive preservation in trans individuals who intend to alter their hormonal system and reproductive organs [ 113 ].

The emergence of egg banking can be considered a different sector in the infertility industry [ 114 ]. The influence of media and the desire for women’s autonomy contributed to the market growth [ 5 , 48 , 49 , 114 – 116 ]; this focuses on the public after 30, a suboptimal age from a clinical point of view, because the quantity and quality of eggs have already decreased considerably [ 108 ].

The ASRM guideline on ART marketing includes a caution to avoid false hopes about delaying procreation, which falls short of what is requested regarding the type and quality of information on most affiliated clinic sites [ 48 , 104 , 117 ].

In parallel, in several situations, the comparative analysis of cost-effectiveness based on direct medical costs demonstrates that the SEF can be financially advantageous in comparison to IVF in older women [ 118 , 119 ]. However, the most efficient/economic strategy for women planning to postpone pregnancy remains uncertain [ 117 ].

The division into clusters was helpful for the identification of topics and do not limit the evaluation of the behavior of the global market, as is the case of the notorious association between moral concerns and national legislation [ 121 ]. Topic modeling proved to be an appropriate tool for detecting terms that allowed us to cluster relevant aspects of this growing market. We were able to identify the size and distribution of this market, as well as list legal, social, and economic aspects, as well as trends and concerns.

Analyzing the ART market is a challengesince many isolated, interdependent, and feedback factors compose it, with cultural and religious associations that cannot be easily evaluated by econometry [ 31 ]. We note that most studies conclude on the need for transnational regulations to solve different issues. We also highlight, the need of more actions in terms of Corporate Social Responsibility, in which the commitment of companies to society occurs based on the practices carried out, going beyond the concept of profits [ 120 ].

We found that, most of the works addresses economic, regulatory, and geographic aspects, and that these topics covered have a synergistic relationship with each other. Two findings gained special attention: a) the potential impact of the formation of conglomerates and mergers on a transnational scale (MAART), considering the certainty about the growing search for reproductive treatments even with legal/social/financial barriers for the final consumer, this has a potential impact on the fragmented pattern of small/medium scale operation, as well as on the CBRC; and b) the lack of health technology assessment (HTA) in reproductive add-ons. Despite technological advances and the insertion of many add-ons over two decades, the success rate remains at around 30% of IVF cycles [ 14 ], especially considering that the ART market devices & consumables were valued at USD 13.75 Billion in 2020 and projected to double by 2028 [ 121 ].

From these findings, it will be possible to establish dynamic and synergistic relationships between the identified topics. This can be used to generate predictive models about the ART market and to point out situations that need to be better understood, such as the low efficiency of IVF cycles. This information can help identify new market niches and increase the availability of technologies and actions for the treatment of infertility.

Limitations of the study

The most significant limitation of this study is the impossibility of exhausting each identified aspect. Also, the generalization of accumulated data causes the loss of local nuances. We would like to create correlation cascades, but we chose not to do so at the risk of creating spurious contexts and escaping the intended purpose of the scoping review.

Supporting information

S1 table. preferred reporting items for systematic reviews and meta-analyses extension for scoping reviews (prisma-scr) checklist..

The elaboration of reviews from the checklist is mandatory for quality studies. We used the specific model for scope reviews according to the model published by Tricco et al . (2018).

https://doi.org/10.1371/journal.pone.0284099.s001

S2 Table. Topics detected by automation tool.

Complete list of terms mined by topic modeling (LDA protocol by Knime). We identified 121 terms covering 7,806 citations.

https://doi.org/10.1371/journal.pone.0284099.s002

S3 Table. The CAGR of the global ART market, according to market reports.

Survey of compound annual growth rate (CAGR) presented in market reports on the subject studied. Note the market growth forecast in all scenarios.

https://doi.org/10.1371/journal.pone.0284099.s003

Acknowledgments

We want to thank the Januário Cicco Maternity Hospital School of the Federal University of Rio Grande do Norte–MEJC/UFRN.

• View Article
• PubMed/NCBI
• Google Scholar
• 4. United Nations. World Fertility Patterns 2015 Data Booklet. UNESA Publication. 2015; 1–30.
• 5. Grand View Research. Assisted Reproductive Technology (ART) Market Size, Share & Trends Analysis Report By Type (IVF, Others), By End Use (Hospitals, Fertility Clinics), By Procedures And Segment Forecasts, 2018–2025. 2018.
• 6. Rockville Research. Global Assisted Reproductive Technology (ART) Market—In Vitro Fertilization, Artificial Insemination and Third Party Assisted ART Market Outlook 2022. 2017.
• 7. Report Buyer. Global infertility testing and treatment market forecast 2019–2027. 2019.
• 8. Boivin J, Gameiro S. Evolution of psychology and counseling in infertility. Fertility and Sterility. Elsevier Inc.; 2015. pp. 251–259. https://doi.org/10.1016/j.fertnstert.2015.05.035 pmid:26092131
• 10. Chambers GM, Adamson GD, Eijkemans MJC. Acceptable cost for the patient and society. Fertility and Sterility. Elsevier Inc.; 2013. pp. 319–327. https://doi.org/10.1016/j.fertnstert.2013.06.017 pmid:23905708
• 11. Rich CW, Domar AD. Addressing the emotional barriers to access to reproductive care. Fertility and Sterility. Elsevier Inc.; 2016. pp. 1124–1127. https://doi.org/10.1016/j.fertnstert.2016.02.017 pmid:27054306
• 12. Global Market Insights, Ugalmugle S, Swain R. Assisted Reproductive Technology Market Size By Procedure (Fresh Donor, Fresh Nondonor, Frozen Donor, Frozen Nondonor, Embryo/Egg Banking), By Technology {In-Vitro Fertilization (IVF) [IVF with Intracytoplasmic Sperm Injection (ICSI), IVF without ICSI], A. 2020.
• 15. Fortune Business Insights. Assisted Reproductive Technology (ART) Market Size, Share & Industry Analysis, By Technique (In-Vitro Fertilization (IVF), Artificial Insemination (AI-IUI), Frozen Embryo Transfer (FET), and Others), By Procedure (Fresh Donor, Fresh Non-donor, Frozen Dono. 2019.
• 16. Market Research Future. Global Assisted Reproductive Technology (ART) Market Report, by Diagnosis (Ovulation Testing, Hysterosalpingography), Type (Ovulation Induction, Artificial Insemination, Surrogacy), End-User (Hospitals & Clinics, IVF Center)–Global Forecast Till 2023. 2020.
• 17. Markets and Markets. Infertility Treatment Market by Product (Equipment, Media, Accessories), Procedure [ART (IVF, ICSI, Surrogate), Insemination, Laparoscopy, Tubal Ligation], Patient Type (Female, Male), End User (Fertility Clinics, Hospitals, Cryobanks)—Forecasts to 2023. 2018.
• 18. Transparency Market Research. Assisted Reproductive Technology Drugs Treatment Market—Global Industry Analysis, Size, Share, Growth, Trends, and Forecast 2018–2026. 2020.
• 30. Strasser MO, Dupree JM. Care Delivery for Male Infertility: The Present and Future. Urologic Clinics of North America. W.B. Saunders; 2020. pp. 193–204. https://doi.org/10.1016/j.ucl.2019.12.006 pmid:32272991
• 36. Vaughn R. IVF Goes IPO | International Fertility Law Group. 2013 [cited 4 Jul 2020]. Available: https://www.iflg.net/ivf-goes-ipo/
• 37. Ware J. First IVF provider to be floated on stock market sees early share price boost—BioNews. In: BioNews [Internet]. 2013 [cited 28 May 2020]. Available: https://www.bionews.org.uk/page_94166
• 38. Markets and Markets. In Vitro Fertilization Market by Product (Imaging System, Incubator, Cabinets, Sperm Separation System, Cryopreservation media, Embryo Culture media), Type of Cycles & End Users (Fertility Clinics and Surgical Centres, Cryobanks)—Global Forecast to 2022. 2017.
• 39. IVI Fertility. IVI arrives at America hand-in-hand with RMANJ and consolidates as the largest group of assisted reproduction in the world. 2017 [cited 4 Jul 2020]. Available: https://ivi-fertility.com/notes/ivi-arrives-at-america-hand-in-hand-with-rmanj-and-consolidates-as-the-largest-group-of-assisted-reproduction-in-the-world/
• 40. Sharanya Pillai. Thomson Medical enters assisted reproduction JV with IVI-RMA Global | The Edge Markets. In: The Edge Markets [Internet]. 2018 [cited 4 Jul 2020]. Available: https://www.theedgemarkets.com/article/thomson-medical-enters-assisted-reproduction-jv-ivirma-global
• 41. Palmer A. Fertility benefits company Progyny files to go public. In: CNBC [Internet]. 2019 [cited 28 May 2020]. Available: https://www.cnbc.com/2019/09/27/fertility-benefits-company-progyny-files-to-go-public.html
• 44. The Needle Starts to Move on Reproductive Health Benefits. [cited 15 Mar 2021]. Available: https://www.mercer.us/our-thinking/healthcare/the-needle-starts-to-move-on-reproductive-health-benefits.html
• 46. Ferraretti AP. Economics of assisted reproductive technology: Data from an international analysis. Fertility and Sterility. Elsevier Inc.; 2014. pp. 47–48. https://doi.org/10.1016/j.fertnstert.2013.09.043 pmid:24188883
• 51. Alviggi C, Esteves SC, Orvieto R, Conforti A, La Marca A, Fischer R, et al. COVID-19 and assisted reproductive technology services: Repercussions for patients and proposal for individualized clinical management. Reproductive Biology and Endocrinology. BioMed Central Ltd.; 2020. p. 45. https://doi.org/10.1186/s12958-020-00605-z pmid:32404170
• 56. Glenn Cohen I. Prohibiting anonymous sperm donation and the child welfare error. Hastings Center Report. John Wiley and Sons Inc.; 2011. pp. 13–14. https://doi.org/10.1002/j.1552-146x.2011.tb00134.x pmid:21980893
• 59. Deomampo D. Transnational Reproduction. NYU Press; 2016.
• 60. Hornstein MD. State of the ART: Assisted Reproductive Technologies in the United States. Reproductive Sciences. SAGE Publications Inc.; 2016. pp. 1630–1633. https://doi.org/10.1177/1933719116667227 pmid:27624310
• 63. Torres G, Shapiro A, Mackey TK. A review of surrogate motherhood regulation in south American countries: Pointing to a need for an international legal framework. BMC Pregnancy and Childbirth. BioMed Central Ltd.; 2019. https://doi.org/10.1186/s12884-019-2182-1 pmid:30691390
• 65. International Federation of Fertility Societies. International Federation of Fertility Societies’ Surveillance (IFFS) 2019: Global Trends in Reproductive Policy and Practice, 8th Edition. Glob Reprod Health. 2019;4: e29–e29. https://doi.org/10.1097/GRH.0000000000000029
• 67. Agigian A. Stratified Reproduction. The Blackwell Encyclopedia of Sociology. Oxford, UK: John Wiley & Sons, Ltd; 2007. https://doi.org/10.1002/9781405165518.wbeoss280
• 69. Gürtin ZB, Inhorn MC. Introduction: Travelling for conception and the global assisted reproduction market. Reproductive BioMedicine Online. Reprod Biomed Online; 2011. pp. 535–537. https://doi.org/10.1016/j.rbmo.2011.08.001 pmid:21962528
• 70. Orion Market Research. Global Assisted Reproductive Technology Market Research and Forecast 2018–2023. 2019.
• 71. Allied Market Research, Shaikh S, Sumant O. IVF Devices and Consumables Market by Product (Instrument, Accessory & Disposable, and Reagent & Media), Technology (Fresh Embryo IVF, Frozen Embryo IVF, and Donor Egg IVF), and End User (Fertility Clinic, Hospital, Surgical Center, and Clinical Research. 2019.
• 72. Transparency Market Research. Assisted Reproduction Technology Devices Market—Global Industry Analysis, Size, Share, Growth, Trends, and Forecast 2016–2024. 2020.
• 73. Grand View Research. Male Infertility Market Size, Share & Trends Analysis Report By Test (DNA Fragmentation Technique, Oxidative Stress Analysis, Microscopic Examination), By Treatment, By Region, And Segment Forecasts, 2020–2027. 2020.
• 78. Allied Market Research, Sumant O, Joshi K. Fertility Services Market by Procedure (IVF with ICSI, IUI, IVF without ICSI, Surrogacy, and Others) and Service (Fresh Non-donor, Frozen Non-donor, Egg & Embryo Banking, Fresh Donor, and Frozen Donor): Global Opportunity Analysis and Industry Forecast, 2. 2019.
• 80. Makinde OA, Makinde OO, Olaleye O, Brown B, Odimegwu CO. Baby factories taint surrogacy in Nigeria. Reproductive BioMedicine Online. Elsevier Ltd; 2016. pp. 6–8. https://doi.org/10.1016/j.rbmo.2015.10.001 pmid:26602942
• 85. Valdez N, Deomampo D. Centering Race and Racism in Reproduction. Medical Anthropology: Cross Cultural Studies in Health and Illness. Taylor and Francis Inc.; 2019. pp. 551–559. https://doi.org/10.1080/01459740.2019.1643855 pmid:31829735
• 86. Riley NE. Stratified Reproduction. Springer, Dordrecht; 2018. pp. 117–138. https://doi.org/10.1007/978-94-024-1290-1_9
• 91. Mamo L. Queering reproduction in transnational bio-economies. Reproductive Biomedicine and Society Online. Elsevier Ltd; 2018. pp. 24–32. https://doi.org/10.1016/j.rbms.2018.10.008 pmid:30723813
• 96. Kooli C. Review of assisted reproduction techniques, laws, and regulations in Muslim countries. Middle East Fertility Society Journal. Springer; 2020. p. 8. https://doi.org/10.1186/s43043-019-0011-0
• 98. Präg P, Mills MC. Assisted Reproductive Technology in Europe: Usage and Regulation in the Context of Cross-Border Reproductive Care. Demographic Research Monographs. Springer, Cham; 2017. pp. 289–309. https://doi.org/10.1007/978-3-319-44667-7_14
• 106. De Geyter C, Wyns C, Mocanu E, De Mouzon J, Calhaz-Jorge C. Data collection systems in ART must follow the pace of change in clinical practice. Human Reproduction. Oxford University Press; 2016. pp. 2160–2163. https://doi.org/10.1093/humrep/dew199 pmid:27591234
• 121. Markets and Markets. Preimplantation Genetic Testing Market by Type (Genetic Diagnosis/Genetic Screening), Technology (NGS, PCR, FISH, CGH, SNP), Products & Services, Application (Aneuploidy, Translocations), End User (Fertility Clinics, Research)—Forecast to 2022. 2017.

Advertisement

Mapping ethical, legal, & social implications (ELSI) of assisted reproductive technologies

• Assisted Reproduction Technologies
• Open access
• Published: 29 June 2023
• Volume 40 , pages 2045–2062, ( 2023 )

Cite this article

You have full access to this open access article

• Ido Alon   ORCID: orcid.org/0000-0001-6603-7496 1 , 2 ,
• Zacharie Chebance 6 ,
• Francesco Alessandro Massucci 4 ,
• Theofano Bounartzi 5 &
• Vardit Ravitsky 2 , 3  

2517 Accesses

3 Citations

3 Altmetric

Explore all metrics

A significant portion of the research on assisted reproductive technologies explores ethical, legal, and social implications. It has an impact on social perceptions, the evolution of norms of clinical practices, regulations and public funding. This paper reviews and maps the geographical distribution to test the hypothesis of geographical concentration and classifies the output by fields and topics.

We queried PubMed, Scopus and the Web of Science for documents published between 1999 and 2019, excluding clinical trials and medical case reports. Documents were analyzed according to their titles, abstracts and keywords and were classified to assisted reproductive fields and by Topic Modeling. We analyzed geographic distribution.

Research output increased nearly tenfold. We show a trend towards decentralization of research, although at a slower rate compared with clinical assisted reproduction research. While the U.S. and the U.K.’s share has dropped, North America and Western Europe are still responsible for more than 70%, while China and Japan had limited participation in the global discussion. Fertility preservation and surrogacy have emerged as the most researched categories, while research about genetics was less prominent.

Conclusions

We call to enrich researchers’ perspectives by addressing local issues in ways that are tailored to local cultural values, social and economic contexts, and differently structured healthcare systems. Researchers from wealthy centers should conduct international research, focusing on less explored regions and topics. More research on financial issues and access is required, especially regarding regions with limited public funding.

Similar content being viewed by others

Mapping international research output within ethical, legal, and social implications (ELSI) of assisted reproductive technologies

Going the Distance

Terminology Transfer in the Field of Bioethics of Assisted Reproductive Technologies

Avoid common mistakes on your manuscript.

Introduction

In the last two decades, the use of Assisted Reproductive Technologies (ART) has been increasing worldwide. The number of yearly performed ART cycles has risen approximately threefold in the United States [ 14 ], fourfold in Europe [ 20 , 22 ], sixfold in Japan [ 29 ] and by at least tenfold in China [ 6 , 39 ]. Meanwhile, the volume of research output in the field has expanded even faster, as indicated by the increasing number of high-quality scientific publications [ 24 ].

A significant portion of the research on Assisted Reproductive Technology (ART) explores the ethical, legal, and social implications (ELSI) of ART, applying humanities and social science methodologies. This research impacts the way individuals and society perceive ART, the evolution of norms of clinical practices, and the way these technologies are being regulated and funded. [ 4 , 15 , 19 , 31 , 35 , 38 ]. Therefore, influenced by cross-cultural differences and value-based perspectives, ART regulations vary significantly between states and countries [ 13 ], which was previously described as legal mosaicism [ 37 ]. Moreover, the complexity and weight of regulations increases steadily, due to the growing use of ART, the technical progress, the dissemination of techniques such as gametes cryopreservation, Preimplantation Genetic Testing (PGT), and the emergence of novel possibilities such as Germline Genetic Modification (GGM), among others [ 4 , 5 ].

Previous reviews analyzed trends in clinical reproductive medical research between 2003 and 2012 [ 2 ], and mapped general trends of ART research between 2005 and 2016 [ 24 ]. This review, focuses on the ELSI—non-clinical literature encompassing ART, excluding clinical trials, medical case reports and laboratory techniques analyses. Our corpus, which was extracted from PubMed, Web of Science and Scopus, includes 7,714 articles concerning ART applications from humanities and social sciences perspectives.

Geographical concentration of ELSI of ART (i.e., in a limited number of countries or centers) may result in findings that are not representative or comprehensive, since local socio-cultural contexts may shape the framing of, and approach towards research questions and analysis. Admittedly, clinical research in ART can also suffer from being non-inclusive and non-representative. When such research is concentrated in a few countries, or even a limited number of centers of excellence, findings may fail to generate data and clinical approaches that are applicable for diverse populations in other locations and contexts. However, such clinical research in ART could still result in some objective outcomes that can be translated into universally relevant protocols and guidelines. Contrariwise, ELSI literature in ART is shaped in a more profound way by socio-cultural contexts and may thus differ in more significant ways between different societies and circumstances.

This review aims to test the hypothesis of geographical concentration and evaluate its scale. We map the literature published between 1999 and 2019 according to country of corresponding authors, taking into consideration a significant portion of international research (16.3% of the corpus), i.e., those cases in which a corresponding author from one country was conducting research about another country. We report trends and shifts in research focus according to predefined ART fields and by topic modeling, and identify gaps and opportunities for researchers. Furthermore, the classification of the literature into fields and topics will enable to conduct various meta-analyses in future research.

We began by establishing inclusion criteria to select articles dealing with ELSI of ART and exclusion criteria to exclude articles dealing primarily with clinical and medical issues, as described in Table 1 .

The corpus was collected from the online databases PubMed, Web of Science (WoS) and Scopus. Since we aimed to analyze titles, abstracts and keywords, we included article that had an abstract in English, regardless of the article’s language.

After running keywords frequency analysis following various preliminary queries, our interdisciplinary team selected three groups of Medical Subject Headings (MeSH)-terms, as shown in Table 2 . Group A included ART terms and Group B included terms indicating disciplines within humanities and social sciences. Group C was formed of terms that are typically used by clinical/scientific articles and rarely by writers of ELSI, in order to exclude irrelevant articles. We carefully aimed to find balance between false-positive (inclusion of articles with medical-clinical nature) and false-negative (exclusion of articles concerning social sciences and humanities).

We used the PubMed API [ 41 ] to query for articles with ‘One MeSH-term from group A’ AND ‘One MeSH-term from group B’ AND ‘Humans (MeSH)’ AND ‘1999–2019’ NOT ‘Any MeSH-term from group C’.

The PubMed query brought up 11,246 results of which 7,003 had a full record of title and abstract in English. Additionally, 159 articles which were queried with no full record from PubMed, were imported from Scopus. In total, 7,162 articles had a full record.

We dropped all abstracts with less than 50 words (259) Footnote 1 ; removed articles if article type included “Clinical Trial”, Controlled Clinical Trial”, “Randomized Controlled Trial” or “Validation Study” (536); and excluded all journal titles containing the words: "animal", "zoo", "plant", "marine", "poultry", "fish", "insect", "wild", "virus", "bacter", "veterin", "botany", "agricult", "avian", "pest", "bug", "maritime", "aquatic", "xenobiotic" (34). We remained with 6,333 articles and extracted a list of keywords including their frequencies within titles, abstracts and keywords.

Using frequency analysis, we selected three groups of keywords (see Table 2 ) with similar definition as described above (for the MeSH-terms), and queried the WoS and Scopus APIs for articles of which the title, abstract or keywords had: ‘One term from group A’ AND ‘One term from group B’ AND ‘1999–2019’ NOT ‘Any term from group C’.

We extracted 14,394 and 12,588 articles from WoS and Scopus, respectively. In addition to the 6,333 extracted from PubMed, 33,315 articles were merged from all three databases. Following the removal of duplicates of titles, abstract and DOI, 17,247 articles remained, of which 14,283 had available title and abstract in English. We repeated the cleaning methods previously applied on the PubMed query (explained above), removed 154 articles due to short abstract (less than 50 words), and 99 due to journal names including the words “animal”, “Zoo” and others (see above). 14,030 articles remained.

Manual cleaning

Two researchers cleaned up the database over the course of six months by analyzing the articles’ titles and abstracts with an emphasis on the following rejection criteria (derived from Table 1 ): Articles dealing exclusively with 1. Animal research. 2. Treatment outcome, unless measured in terms of population and socio-economic characteristics (i.e., national registries). 3. Clinical policies on a local level (in contrast to national/regional policies). 4. Clinical outcome and performance of technologies and protocols. 5. Processes within hospitals and clinics. 6. Prenatal testing and selection. 7. Therapeutics (non-ART) uses of stem-cell research. 8. Clinical trials or reports in psychiatry/psychology with no ELSI or impact on patients’ decision making.

We removed 6,316 articles that did not meet our criteria, leaving a database of 7,714 relevant articles of full record. Among these 7,714 articles, 1,184 were non-English articles with title and abstract in English, allowing for analysis and classification. These non-English articles made up 15.34% of the database.

Classification

The abstracts were processed by merging the title and abstract into one string (“code”). We harmonized the “codes” to British English, tokenized Footnote 2 each sentence into a list of words, removed punctuations and unnecessary characters and replaced upper-cases with lower-cases. Stop words (‘a’, ‘the’, ‘is’, ‘are’ etc.) were removed Footnote 3 and the text was lemmatized (inflectional endings were removed, leaving the root words). Footnote 4 We formed a list of terms to be replaced with acronyms or abbreviations in order to unify the text and allowing us to identify repetitions. For examples: ‘in vitro fertilization’ was replaced with ‘ivf’, ‘oocytes cryopreservation’ and ‘eggs cryopreservation’ with ‘eggscryopreservation’.

Next, we extracted a list of terms by the frequency of “codes” in which they appear and divided the most frequent technical-medical terms into ten ART fields (see Appendix 1 ): 1. Egg Donation; 2. Sperm Donation; 3. Embryo Donation; 4. Surrogacy; 5. Fertility Preservation; 6. Stem-cell Research; 7. PGT; 8. Genetic Modification; 9. MRT; 10. Assisted Insemination. An article was assigned into an ART field (or several) if one term, associated to that field, was found in its title, abstract or keywords. The articles which included none of these terms were assigned by two researchers reading their titles and abstracts. The remaining articles (2,267/7,714) were appointed as “General”.

Subsequently, for each publication, all obtainable metadata was extracted from WoS, Scopus and PubMed, merged and unified under one template. Finally, we cleaned the abstracts from all those technical-medical terms that were used to define the 10 ART fields and then: 1. Uploaded the database to the VOSviewer software tool for constructing and visualizing bibliometric networks. 2. Applied Topic Modeling (TM) via Latent Dirichlet Allocation (LDA) [ 9. , 32. ] to all “codes”. To do so, we've built a corpus dictionary using gensim Footnote 5 open-source library. Footnote 6 The LDA method assumes that the observed distribution of words in a textual corpus is determined by a statistical model that fixes both a word-topic and an article-topic association [ 24 ].

We defined the number of topics by computing the coherence score as a function of the number of topics Footnote 7 and by assessing the results in relation to the VOSviewer analysis. The results of the LDA algorithm consist in a list of topics, and in the weighted relations (0 to 1) between each article and each topic. Every topic is a list of characterizing words, and each article may be connected to more than one single topic. Thus, as seen in Fig.  1 , at least 5 topics should be identified. According to our analysis, we defined 6 topics as the most accurate solution. Each article was associated to those topics by considering a weighted relation of more than 0.333334 so that each article could be associated to not more than two topics. As a result, only 161 articles remained unassociated.

Coherence Score as a function of the number of topics

In addition to the VOSviewer tool, we conducted an analysis using Microsoft Excel Spreadsheet based on year of publication, countries of corresponding authors and countries mentioned in codes as an indicator for the focus of the article. For the last, we searched the codes (abstracts, titles and keywords) using a list of countries, cities (with no duplications), and nationalities. Footnote 8 ART Fields, topics identified by the TM, regions, Footnote 9 and income level. Footnote 10 We used the IBM SPSS Statistics tool to test for Spearman’s correlations between the ART fields and the topics (Appendix 2 ). Those correlations were used for two purposes. First, to group ART fields in order to simplify the results, i.e., [Egg (C1) and Sperm donation (C2)], and [Stem-cell Research (C6) PGT (C7) and Genetic Modification (C9)]. Second, to verify a relationship between ART fields and topics.

Between 1999 and 2019, the global research output in the field grew about 12% annually. Therefore, the annual output increased almost tenfold (× 9.75), from 72 publications in 1999 to 702 in 2019. We begin by analyzing research areas on two axes, the first includes eleven ART fields (10 technical fields plus General) and six Topics (clusters) identified by both TM and the VOSviewer analysis. Additionally, we present a geographical analysis of the database.

Areas of research

Some articles were assigned, to more than a single ART field or a single topic. For this reason, in Figs.  2 and 3 (as well as, Figs. 5 , 8 and 9 ), the total sums-up to more than 100%. The three ‘Donation’ fields had a variable trend during the two decades which ended in an overall decrease from 44% (1999) to 30% (2019) of the total. The share of ‘Assisted insemination’ has decreased steadily while both ‘Fertility Preservation’ and ‘Surrogacy’ are the two emerging fields throughout this period, particularly in the second decade. In the last three years each of these two fields was engaged by more than 20% of the literature. Furthermore, ‘Stem-cells’, ‘PGT’ and ‘Genetic Modification’ were the most explored fields at the beginning of the previous decade, occupying 68% of all publications in 2005. During the second decade, interest in the genetics of ART has moderated while the field of ‘MRT’ has emerged.

Based on Spearman correlations (Appendix 2 ) we grouped the fields ‘Egg Donation’ and ‘Sperm Donation’ (into ‘Egg & Sperm Donation’), as well as, the fields Stem-cell research, PGT and Genetic Modification (into ‘Genetics’), in order to simplify result presentation in the forthcoming analysis.

We used the VOSviewer software tool to find co-occurrence links between terms. Footnote 11 In Fig.  4 , the size of the label and the frame of an item is determined by the weight (number of occurrences) of the item. Thicker Lines between items represent stronger links [ 30 ]. The colors represent 5 clusters. We invite the reader to consult the interactive version of this figure, available online at: https://app.vosviewer.com/?json=https://drive.google.com/uc?id=1iQXv7oEQ6RE7-wz-578G00ULsLfkL3ao

Additionally, the TM raised 6 topics which are characterized by lists of words of which we displayed the first 10 for each topic, as seen in Table 3 (full output in Appendix 3 ). We assigned these topics to the clusters identified by the VOSviewer analysis (The Red was the largest cluster and was assigned with two topics):

Co-occurrence of keywords

We may clearly see a decrease in the share of topic ‘Bioethics’ with increases in the shares of both ‘Law and Policy’ and ‘Attitudes and Knowledge’. The share of ‘Psychology’ decreased in the first decade but recovered during the second. We should remind that, due to the database selection, ‘Psychology’ includes only articles with ELSI or impact on patients’ decision making and not all psychological research concerning ART.

As seen on Fig.  5 and according to Spearman’s correlations presented in Appendix 2 , we note that ‘Bioethics’ is strongly focused on the ‘Genetics’ field, though this focus has decreased with time, from 73% in the first period to around 53% in the last. The field ‘Egg & Sperm Donation’ has gained share under the topics ‘Bioethics’ and ‘Law and Policy’. It was nevertheless the focus of More than 45% of the topic ‘Family and Sociology’, while a relatively high but decreasing share of this topic dealt with ‘Assisted Insemination’. Concerning the two emerging fields, 'Fertility Preservation' was most strongly covered under the topics 'Cost and Outcome’ and 'Attitudes and Knowledge', while ‘Surrogacy’ was highly under the focus of the topics ‘Law and Policy’ and ‘Family and Sociology’. The decreasing total percentages indicates that research is becoming more specialized, i.e., for all topics except for ‘Bioethics’, with time, the average number of fields related to a single article is decreasing. The topics ‘Cost and Outcome’ and ‘Family and Sociology’ are more multi-field than the others.

ART fields in topics

Geographic analysis

Figures  6 and 7 present the shares of publications per regions and leading countries Footnote 12 respectively. In 1999, the five leading regions Northern America, Northern, Southern and Western Europe, as well as Australia-New Zealand were responsible for 90% of the world publications (according to corresponding authors). Nevertheless, in 2019 the share of the five leading regions decreased to less than 76.5%. The Northern American share decreased from 40 to 26%, while Southern Europe’s share increased from 7 to 20% and this region became the second most engaged.

Regions (by Corresponding Author)

Countries of corresponding author. The leaders and followers are ordered by the total publications of 1999–2019 while the emergent countries are ordered by the publications of 2019

The two leaders, the US and the UK, fell from 51% in 1999 to 31% in 2019, although still holding a significant share of global publications. In the middle part of Fig.  7 , we may notice that among the eleven followers, the shares of Australia, Canada, New Zealand, the Netherlands, Germany and Denmark, all decreased significantly, while the others have been relatively stable.

On the bottom part of Fig.  7 , we observe the emergence of Spain (the fourth largest ART user in the world) and Italy, as well as the awakening of two world’s leaders in terms of ART cycles, China and Japan, which during the last decade increasingly contributed to the global ART literature.

Of the 7,714 articles included in this review, 3,743 (48.5%) were identified as dealing with specific countries, while 1,260 (16.3%) were international research, in which the corresponding author was based in one country while conducting research about another.

Table 4 demonstrates the 20 leading countries in international research alongside the 23 most popular research subjects. All of the leading countries in the field of ELSI of ART were also engaged in international research, despite their research being largely an exchange between the leaders and the followers, with the exception of India, that attracted a large number of foreign researchers (121 articles), particularly in the field of surrogacy (85 articles).

Wealthy nations (according to the World Bank’s level of income) were much more engaged in research about ELSI of ART. Hence, 89% of all publications had a corresponding author from a high-income country, only 8% from a high-middle income country, and 3% from a low-middle income country (there were only 14 publications from low-income countries throughout the entire period). High-income countries had more focus on ‘Egg & Sperm Donation’ and on ‘Genetics’ while high-middle and low-middle income countries had more focus on ‘Surrogacy’. The most engaged among the high-middle and low-middle income countries were Brazil, Iran, India, Turkey, China, South Africa, Russia, Romania, Nigeria, and Mexico.

There were no remarkable specializations of certain regions in certain topics, as seen in Fig.  8 . For all topics, the US leads soundly, followed by the UK. The following countries in all topics are Australia (with the exception of ‘Cost and Outcome’) and France (with the exception of ‘Psychology’). The relatively high share of the topic ‘Psychology’ among Western and Southern Asian publications could be explained by the attention-grabbing focus of Iran in this topic (53 out of its 87 publications), which makes it the fourth most publishing country in this topic during those two decades. Moreover, for all topics exccept ‘Psychology’ (66%), the 10 most publishing countries are responsible for 70–80% of the global publications.

Topics by region

There were some significant trends in ART fields (Table 5 and Fig.  9 ). The US was the leader in almost all fields during the two decades and was always followed by the UK, that led only in publications concerning ‘MRT’ (with 34% of the total). In this emerging field, Canada was third, and the ten leading countries were responsible for 82% of the global publications. In the field of ‘Egg & Sperm Donation’, Italy emerged strongly, reaching the top five at the end of the period. In the minor field ‘Embryo donation’, German authors were dominant, publishing 5% of the global amount. Iran was also strong in this field and was 6 th with 17 publications (14 of them in the second decade). Research concerning ‘Surrogacy’ was the most decentralized as “only” 69% were published by the top 10 countries. In this field Spain and Italy emerged to become major publishers. Under the ‘Genetics’ fields, particularly concerning PGT, Germany was the third strongest publisher. However, its engagement was felt more at the first decade, whereas Australia was stronger at the second decade. In all the three fields, ‘Egg &S perm Donation’, ‘Fertility Preservation’ and ‘Genetics’, 75% of global publications were published by the top 10 countries.

ART fields by regions

To summarize (Fig.  9 ), Northern, Western and southern Europe were slightly more engaged in the research about ‘Egg & Sperm Donation’, while Northern America, as well as Australia and New Zealand were more focused on ‘Genetics’ and ‘Fertility Preservation’. The regions which focused more on Surrogacy were Southern Europe, Western-Southern Asia, and Eastern Europe and Central Asia.

We conducted this review to map out the geographical distribution of ELSI of ART research output published between 1999 and 2019, according to country of corresponding authors. We also sought to explore research themes in order to identify concentration of research attention and gaps. Geographic/regional and topical centralization of ELSI studies on ART produces research gaps. Such studies are intended, amongst other things, to inform state and professional regulations that largely differ between nations due to local legal, social and economic contexts, cultural values, and diverse structures of healthcare systems. Diversity of research locations and topics is important to ensure attention is given to all relevant ELSI themes in various cultural and socio-economic contexts. Lack of diversity creates blind spots that are problematic for patients, families, communities, clinicians, and decision-makers.

In these 21 years, ELSI of ART research output increased nearly tenfold, following an average yearly growth rate of 12%. This growth was much higher compared with that of the entire literature on ART (including all ELSI and clinical/scientific articles), which had an overall increase of only 20% between 2005 and 2015, i.e., an average yearly growth rate of merely 1.7% [ 24 ]. Growth in the output of ELSI of ART was also much higher than the 4% annual growth rate of global scientific output between 2008 and 2018, as reported by the US National Center for Science and Engineering Statistics [ 42 ].

The need for de-centralized and diverse ELSI research on ART

Overall, our findings show a clear trend towards decentralization of ELSI of ART research, but they also show that much work remains to be done in terms of further diversification of research regions and themes. The Global scientific output has been centralized for many years within the U.S. and the U.K. as dominant centers of academic research. Our early hypothesis was thus that publications on ELSI of ART would be centralized in a limited number of countries. However, scientific research has become more globalized in recent years, with the share of the U.S. and the U.K. declining. In particular, China’s contribution to research output has grown remarkably, both regarding general science and in relation to ART research. The previous centralized landscape has been replaced by a more diverse academic world, centered in North America, Europe, and Asia–Pacific [ 18 , 40 , 42 ].

Our review shows that research on ELSI of ART has indeed followed this trend over the past two decades, but to a considerably lesser extent compared with scientific research or clinical ART research. In ELSI of ART publications, while the U.S. and the U.K.’s share has dropped from more than 50% in 1999 to around 30% in 2019, North America and Western Europe (including north and south) are still responsible for more than 70% of the output, and adding Australia and New Zealand increases the number to 80% (Fig.  6 ). While international research helped mitigate the concentration of research in a limited number of countries, a large part of it was conducted across countries in the leading regions (Table 4 ).

These figures may raise concerns if we consider that China and Japan are currently the world leaders in annual ART cycles, and that China alone performs more cycles than the European Union, the U.K. and Russia combined [ 6 , 21 , 39 ]. Indeed, regarding clinical ART research output, China has become a world leader in recent years, third after the U.S. and the U.K. [ 24 ]. However, regarding research on ELSI of ART, Eastern and Southeastern Asia were responsible for merely 4–6% of the global output in the end of the studied period, between 2016 and 2019, still far behind North America and the European regions.

There are several possible explanations for this discrepancy. One might be that language barriers are reducing research output in this area. However, other countries that face language barriers, such as Spain (a world leader in ART) and Italy, have been increasingly contributing to ELSI of ART, showing that such barriers can be overcome. Further, it is possible that some or much of the work in this area is published in local languages, such as Mandarin or Japanese, a question that can be empirically assessed by future research. In our corpus, out of the 1,184 non-English articles (with English abstracts) only 11 articles were written in Mandarin and 12 in Japanese, and most (88%) were in German, Spanish, French, Italian and Portuguese.

Another possible explanation might be cultural differences. ELSI issues might be framed differently in countries with non-Western value systems, affecting the type and quantity of research in this area. Cultural framing may even influence the perception or categorization of certain aspects of ART as ‘issues’ that ought to be researched. What may be seen in one socio-cultural context as a problem that needs to be addressed, might be seen in another as a mere social fact that does not raise concerns worthy of investigation.

In China, many of the ART techniques often discussed by western literature are restricted by law. For example, surrogacy, egg donations (except for sharing between patients), embryo donation (except for research purposes), and fertility preservation (except for married couples and for medical reasons) are prohibited. Other practices, such as sperm donation, are marginalized by social norms. In general, ART is allowed only for “healthy”, officially married couples, and the eligibility for LGBTQ individuals is barely even discussed [ 1 , 27 , 33. ]. In Japan, many of these issues were, at least until recently, unregulated and marginalized. Overall, ART practices usually associated wth ELSI research are much less common in Japan, compared with other ART leading countries [ 28 ]. Thus, different context and framings may have an impact on the quantity of research that is conducted and published in various regions.

In light of differences in cultural values and hence in regulations between nations, cross-border reproductive care provides opportunities to overcome legal barriers for those who can afford it, but can create inequalities, tensions and frustration [ 26 ], and increase risks to women and children involved. Consequently, there are often attempts to discuss ART regulations at the international level, particularly concerning exceedingly controversial issues such as genetic modification and surrogacy. Often, ELSI studies are elaborated as an ongoing international discussion between scholars and experts around the globe. The key to a successful global environment of regulatory collaboration, at least regarding some critical ART applications, is the production of more comprehensive and diverse research. We call to enrich researchers’ perspectives with diverse local expertise, addressing local issues in ways that are tailored to local cultural values, social and economic contexts, and differently structured healthcare systems. This need for diverse research on ELSI of ART is particularly salient regarding countries and regions with a high use of ART.

Trends in research fields and topics

Some of the trends our findings have identified regarding ART fields and research topics are unsurprising. In most countries with developed ART services, egg, sperm and embryo donations concern a very large share of ART cycles and raise frequent socio-ethical tensions. The share of these fields in the literature fluctuated throughout the last two decades and despite declining by 2019, it is hard to determine a downtrend (Fig.  2 ). In European regions, where both an aging population and delayed childbearing are particularly prominent, we observed a significant and mostly increasing proportion of ELSI research about egg, sperm, and embryo donations (Fig.  9 ). The heightened interest in these areas is probably closely tied to the socio-ethical questions they provoke, such as the definition of parenthood, parental rights and responsibilities, as well as the rights of children to be informed of their genetic origins. Consequently, these matters have led to a diverse range of regulations and clinical norms across the European continent. It's noteworthy that the largest shares of ELSI research on egg and sperm donations occur in Northern and Western Europe, regions where the most stringent legal restrictions are in place. The emergence of research on ELSI of fertility preservation and MRT is also understandable, as both technologies matured and became practical during the study period.

Some specific findings can also be explained by considering local factors. For example, the UK led in publications concerning MRT (34% of the total papers on this topic), possibly since it was the first country in the world to regulate the practice following a national consultation [ 16 ]. In this emerging field, Canada was third, which can be explained by the opposite reason – MRT is legally banned there and conducting it would constitute a federal criminal offence, which led to a discussion of this barrier for Canadians [ 16 ]. In the field ‘Egg & Sperm Donation’, Italy emerged as a strong contributor (2013–2019), possibly effected by the 2014 rule of the Italian Constitutional Court which overturned the ban on gamete donation [ 11 ]. The focus on ELSI of fertility preservation in the U.S. could be driven by the advent of novel techniques for cancer patients, notably children, and the vigorous socio-ethical debate surrounding social egg freezing. This trend, largely unsupported by evidence, is propelled by commercial incentives, media coverage, and corporate benefits, yet is accompanied by a dearth of adequate information. The societal implications of advanced maternal age are being addressed, often controversially, by this expensive, unproven method, which tends to shift the burden onto individual women [ 7 , 34 ].

In the field of ‘Embryo donation’, German authors were dominant, which can possibly be explained by some unique German values, attaching the status of person to the human embryo [ 12 ] and making the option of donating embryos for reproduction preferable to other options such as destruction or donation for research and training. Similar cultural explanations can clarify why Germany was the third strongest publisher on the PGT, and why its engagement was felt more at the first decade (1999–2009), a time during which PGT was legally banned there for historical and cultural reasons [ 10 ] and during which much bioethics debate surrounded this ban.

As noted, research on ‘Surrogacy’ was the most decentralized (69% of papers published by the top 10 countries), with Spain and Italy emerging as major publishers, in addition to Austria (Table 5 ). This can possibly be explained by the ban on surrogacy in these countries raising discussions on this matter. The focus on ELSI of surrogacy in Canada can be attributed to the federal legislation prohibiting surrogate payment. This sparked a prolonged debate on the permissibility of specific reimbursements and compensations, which was finally resolved after many years, filling a widely criticized policy void [ 8 ]. In France, the Bioethics Act of 1994 placed strict limitations on surrogacy [ 23 ] resulting in substantial ELSI inquiries. This act, effectively stifling surrogacy agreements, resulted in complex legal predicaments for couples seeking surrogacy abroad. These included complications in acknowledging parentage and securing citizenship for their children, with some children only gaining recognition as citizens upon reaching adulthood.

Moreover, a large share (31%) of international research on surrogacy were focused on India (and 5% on Thailand). This could be partially explained by the ethical tensions surrounding the use of surrogates from low-income countries, especially by couples from high-income countries [ 36 ] and especially when access to surrogacy is limited in the commissioning couple’s country of origins.

However, the decline in the share of ELSI research on stem cell research, PGT and genetic modification is quite unexpected, considering the improvement in these technologies throughout the study period, the increased use of preimplantation genetic screening, and the breakthroughs in CRISPR technologies since 2012. Certainly, a decrease in the share of the ‘Genetics’ field is not a decline in absolute terms, i.e., the number of articles addressing ‘Genetics’ increased in absolute terms over the last two decades, but less than other fields. Yet, with the frequent addition of PGT (mainly PGT-A) to ART [ 14 ], the advancement in germline genetic modification, one would expect an even faster growth of the discussion of socio-ethical tensions concerning these technologies. This decline could be attributed to several factors. Between 1999 and 2005, breakthroughs such as cloning and the completion of the human genome project stimulated significant attention. However, after 2005, the excitement around these technologies may have begun to stabilized, and the establishment of regulatory frameworks could have prompted a shift in research focus. Additionally, as researchers became more aware of recurring ethical and procedural considerations, they might have refrained from duplicating studies. It is also important to remember that influential contributions in these areas often come from grey literature, such as reports from leading organizations, not just from peer-reviewed academic papers.

Observing the topics raised by LDA (Fig.  4 ), due to the correlation between the topic Bioethics and the fields Stem-Cells, PGT and Genetic Modification, we may associate the drop in the portion of bioethics with the decline in the portion of these genetic fields (Appendix 2 ). We also identify a significant increase in the portion of ‘Attitudes and Knowledge’ which could be related to the growing interest in public understanding of both infertility and ART. ‘Law and Policy’, which include many comparative studies, became a more popular type of research throughout the period in all regions. Conversely, the share of ‘Cost and Outcome’ declined towards the end of the period, and we would have expected more publications of this topic, mainly in less wealthy regions where access to ART remains a critical ethical issue. Considering that even in countries with a fairly robust public health system, ART remains at least in part a private service [ 3 ], and that in many places around the world a large share of the population struggles to fund and access ART, we would suggest that there is a challenging gap in the literature concerning cost and the burden of ART funding on families. This research direction would be particularly interesting in the U.S., China, and in LMICs. It is also important to explore barriers to access in light of local cultural norms related to reproduction and the stigma of infertility, which still constitutes a significant psycho-social burden in several regions.

Study limitations and future research directions

Our analysis is not exempt from some methodological limitations. First, our selection and cleaning process had a certain level of subjectivity. The selection criteria were complex and included terms selected by the authors. During the second search cycle focusing on the Web of Science and Scopus, we grouped keywords based on frequency analysis of the articles collected from PubMed. While the initial group addressing Assisted Reproductive Technologies (ART) was comprehensive, the subsequent group intended to encapsulate ELSI research might have benefited from an expansion to include additional terms such as 'anthropology,' 'kinship,' 'cultural beliefs,' and 'social structure.' The omission of these terms potentially explains why anthropology as a topic was not captured in our resulting topic modeling. In our cleaning process, many articles were “manually” removed by analyzing their abstracts, and both false-positives and false-negatives were possible. Second, to facilitate our analysis, we limited our selection to articles with titles and abstracts available in English, therefore, inadvertently excluded some articles published in local or regional journals. Third, in the classification process, the ART fields were defined according to keywords selection, while the occurrence of one term in the abstract indicated an affiliation to an ART field which could lead to some mistakes. Fourth, associating an article to a country based solely on the affiliation of the corresponding author reflects the funding source. Alternatively, focusing on the research topic, as well as co-authors, could indicate that other countries were involved or studied, as shown in a preliminary analysis of international research (Table 4 ). We further explore this in another article.

Finally, topic modeling by LDA has some limitations [ 25 ]. The algorithm assumes a certain probabilistic distribution behind the word/article association, which may not hold in reality. Moreover, there is no "standard" or objective way of fixing the number of topics, which remains a free parameter, adjusted in accordance with different metrics. Consequently, due to the number of topics one eventually selects, some smaller topics may remain hidden. It is also important to remind that in topic modeling the categorization of articles into topics is determined solely by the specific terminology found within each article. Despite these limitations, our findings are based on a comprehensive corpus collected from three major databases, a relatively strong LDA and an additional method of categorization.

In terms of developing our mapping strategies, it would be appealing to analyze international research by identifying studies that were conducted by a corresponding author of one country, but were in fact co-authored by researchers from other countries or were dealing with a different country as a topic. In this paper, we present a preliminary analysis of international research in the ELSI of ART. Our database allows to identify the focus of international research on specific fields and topics, as well as trends over time. Moreover, the division of the database into ART fields and topics would enable various meta-analyses by extracting a single ART field (or topic) and presenting its distributions of study designs, key issues, research questions and outcomes, according to geographic location and timeline.

In this paper we mapped the ELSI of ART literature over 21 years. We have shown geographic centralization in research, based on corresponding author, which reflects an unequal distribution of research across nations. While it is hard to expect a redistribution of research funds, and although our search parameters might not have captured all pertinent international research in the field, it is required and realistic to encourage researchers from wealthy academic centers to conduct more international research and focus on less wealthy or less explored regions and topics. We conclude that the literature on the ELSI of ART is mainly produced by North America, Western Europe and Australia, although it has been slowly decentralized. In the last few years, the leading emerging fields in ELSI of ART researchers have focused on were surrogacy and fertility preservation, with a strong focus on law and policy. Research about patients’ attitudes and knowledge has been rising, while more research on financial issues and access to treatment is required, especially regarding regions with limited or incomplete public funding of ART.

Data Availability

The data that support the findings of this study are available on request from the corresponding author, [IA].

When applying Latent Dirichlet Allocation (LDA) for Topic Modeling, shorter abstracts typically provide insufficient data to yield reliable results.

To lowerize and tokenize we used: https://tedboy.github.io/nlps/generated/generated/gensim.utils.simple_preprocess.html

We used a combination of open sources libraries including: nltk.corpus.stopwords.words('english') and https://github.com/seinecle/Stopwords/blob/master/src/main/java/net/clementlevallois/stopwords/resources/en.txt

First, we have formed bigram models using gensim open source library. Then, to lemmatized we use the open-source library spacy. ( https://spacy.io/usage/models ).

“gensim.corpora.Dictionary”.

I.e., transforming words/sentences into vectors (with 1 and 0): word: id mapping to apply TM afterwards [ 17. ].

https://radimrehurek.com/gensim/models/coherencemodel.html for TM using gensim.models.LdaMulticore with the following parameters: gensim.models.LdaMulticore(corpus = corpus, id2word = id2word, num_topics = k, random_state = 100, chunksize = 100, passes = 10, per_word_topics = True), k number of topics ranging from 2 to 15.

For example, “Portug” to capture Portugal and Portuguese, “Spain” and “Spani”, “Poland” and “Polish”. Also, cities names with double meaning (i.e., Male, Huntington) were carefully removed.

According to the United Nations Geoscheme: https://unstats.un.org/unsd/methodology/m49/ (some regions were merged due to proximity and low number of publications, i.e., Latin Americas and the Caribbean, Western and Southern Asia, Eastern Europe and Central Asia, Eastern and South-Eastern Asia).

According to the historical classification by the World Bank (per year of publication):

https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-lending-groups

With a minimum cluster size = 3, and a minimum number of occurrences of a keyword = 20.

Russia was not among the 23 leading countries in the all period but only in the last years.

Ahuja KK. China’s model of egg donation is a policy lesson for Britain. Reprod Biomed Online. 2012;24:257–60. https://doi.org/10.1016/j.rbmo.2012.01.002 .

Article   PubMed   Google Scholar  

Aleixandre-Benavent R, Simon C, Fauser BC. Trends in clinical reproductive medicine research: 10 years of growth. Fertil Steril. 2015;104(1):131-7.e5. https://doi.org/10.1016/j.fertnstert.2015.03.025 .

Alon I, Pinilla J. Assisted reproduction in Spain, outcome and socioeconomic determinants of access. Int J Equity Health. 2021;20:156. https://doi.org/10.1186/s12939-021-01438-x .

Article   PubMed   PubMed Central   Google Scholar  

Alon I, Guimon J, Urbanos-Garrido R. Regulatory responses to assisted reproductive technology: a comparative analysis of Spain and Israel. J Assist Reprod Genet. 2019;36:1665–168. https://doi.org/10.1007/s10815-019-01525-7 .

Alon I, Urbanos-Garrido R, Guimon J. Regulating reproductive genetic services: dealing with spiral-shaped. J Assist Reprod Genet. 2021;38:305–17. https://doi.org/10.1007/s10815-020-02017-9 .

Bai, et al. Assisted reproductive technology service availability, efficacy and safety in mainland China: 2016. Human Reprod. 2020;35(2):446–52. https://doi.org/10.1093/humrep/dez245 .

Article   CAS   Google Scholar  

Baldwin K, Culley L, Hudson N, Mitchell H. Reproductive technology and the life course: Current debates and research in social egg freezing. Human Fertility. 2014;17(3):170–9. https://doi.org/10.3109/14647273.2014.939723 .

Baylis F, Downie J. The tale of assisted human reproduction Canada: a tragedy in five acts. Can J Women Law. 2013;25(2):183–201 ( https://ssrn.com/abstract=3491814 ).

Article   Google Scholar  

Blei D, Ng A, Jordan M. Latent dirichlet allocation. J Mach Learn Res. 2003;3:993–1022. ( https://www.jmlr.org/papers/volume3/blei03a/blei03a.pdf ).

Google Scholar  

Bock von Wülfingen B. Contested change: how Germany came to allow PGD. Reprod Biomed Soc Online. 2016. https://doi.org/10.1016/j.rbms.2016.11.002 .

Boggio A. The Legalisation of Gamete Donation in Italy. Eur J Health Law. 2017;24(1):85–104. https://doi.org/10.1163/15718093-12341409 .

Bowman-Smart H, Wiesemann C, Horn R. Non-invasive prenatal testing in Germany: a unique ethical and policy landscape. Eur J Hum Genet. 2022. https://doi.org/10.1038/s41431-022-01256-x .

Calhaz-Jorge, et al. Survey on ART and IUI: legislation, regulation, funding and registries in European countries: the European IVF-monitoring Consortium (EIM) for the European Society of Human Reproduction and Embryology (ESHRE). Human Reprod Open. 2020;2020(1):hoz044. https://doi.org/10.1093/hropen/hoz044 .

CDC (2020) Assisted Reproductive Technology - National Summary Report 2018. Atlanta, US,: Centers for Disease Control and Prevention, American Society for Reproductive Medicine, https://www.cdc.gov/art/pdf/2018-report/ART-2018-Clinic-Report-Full.pdf

Chambers G, Adamson D, Eijkemans M. Acceptable cost for the patient and society. Fertil Steril. 2013;100(2):319–27. https://doi.org/10.1016/j.fertnstert.2013.06.017 .

Cohen IG, Adashi EY, Gerke S, Ravitsky V. The Regulation of Mitochondrial Replacement Techniques Around the World. Annu Rev Genomics Hum Genet. 2020;21:565–86. https://doi.org/10.1146/annurev-genom-111119-101815 .

Article   CAS   PubMed   Google Scholar  

Construct word<->id mappings (n.d.) Retrieved from Gensim: https://radimrehurek.com/gensim/corpora/dictionary.html

Csomós G. A spatial scientometric analysis of the publication outputof. J Informetrics. 2018;12:547–66. https://doi.org/10.1016/j.joi.2018.05.003 .

Doetsch JN, Marques S, Krafft T, Barros H. Impact of macro-socioeconomic determinants on sustainable perinatal health care in Portugal: a qualitative study on the opinion of healthcare professionals and experts. BMC Public Health. 2021;21(1):210. https://doi.org/10.1186/s12889-021-10194-0 .

ESHRE (2017) Trends over 15 years in ART in Europe: an analysis of 6 million cycles. Human Reproduction Open 1–10. https://doi.org/10.1093/hropen/hox012

ESHRE. Assisted reproductive technology in Europe, 2016: results generated from European registers by ESHRE. Hum Reprod. 2020. https://doi.org/10.1093/hropen/hoaa032 .

ESHRE (2021) ART in Europe, 2017: results generated from European registreis by ESHRE. Human Reprod Open. 1–17. https://doi.org/10.1093/hropen/hoab026

Fournier V, Spranzi M. The French bioethics debate: norms, values and practices. Med, Health Care Philos. 2013;16:41–4. https://doi.org/10.1007/s11019-012-9428-2

García D, Massucci FA, Mosca A, Ràfols I, Rodríguez A, Vassen R. Mapping Research in Assisted Reproduction Worldwide. Reprod Biomed Online. 2020;40(1):P71-81. https://doi.org/10.1016/j.rbmo.2019.10.013 .

Gerlach M, Peixoto T, and Altmann E (2018) A network approach to topic models. Sci Adv 4(7). https://doi.org/10.1126/sciadv.aaq1360 . 

Gianaroli L, Crivello AM, Stanghellini I, Ferraretti AP. Reiterative changes in the Italian regulation on IVF: the effect on PGD patients’ reproductive decisions Reproductive BioMedicine Online. 2014;28:125–132. https://doi.org/10.1016/j.rbmo.2013.08.014 .

Heng B. Stringent regulation of oocyte donation in China. Hum Reprod. 2009;24(1):14–6. https://doi.org/10.1093/humrep/den352 .

Hibino Y, Allan S. Absence of laws regarding sperm and oocyte donation in Japan and the impacts on donors, parents, and the people born as a result. Reprod Med Biol. 2020;19:295–8. https://doi.org/10.1002/rmb2.12329 .

Ishihara O, Jwa S, Kuwahara A, et al. Assisted reproductive technology in Japan: A summary report for 2018 by the Ethics Committee of the Japan Society of Obstetrics and Gynecology. Reprod Med Biol. 2021;20:3–12. https://doi.org/10.1002/rmb2.12358 .

Jan van Eck N and Waltman L (2021) VOSviewer Manual, version 1.6.17. CWTS, Universiteit Leiden.  https://www.vosviewer.com/documentation/Manual_VOSviewer_1.6.17.pdf .

Johnson M, Petersen K. Public Interest or Public Meddling? Towards a Subjective Framework for the Regulation of Assisted Reproduction Technologies. Hum Reprod. 2008;3:716–28. https://doi.org/10.1093/humrep/dem387 .

Kapadia S (2019) Evaluate Topic Models: Latent Dirichlet Allocation (LDA) . Retrieved from towardsdatascience.com: https://towardsdatascience.com/evaluate-topic-model-in-python-latent-dirichlet-allocation-lda-7d57484bb5d0

Liao J, Dessein B, Pennings G. The ethical debate on donor insemination in China. Reprod Biomed Online. 2010;20:895–902. https://doi.org/10.1016/j.rbmo.2010.01.014 .

Martin L. Anticipating infertility egg freezing, genetic preservation, and risk. Gend Soc. 2010;23:4. https://doi.org/10.1177/0891243210377172 .

Mladovsky P, Sorenson C. Public financing of ivf: a review of policy rationales. Health Care Anal. 2010;18:113–28.

Pande A. Transnational commercial surrogacy in India: gifts for global sisters. Reprod Biomed Online. 2011. https://doi.org/10.1016/j.rbmo.2011.07.007 .

Pennings G. International Evolution of Legislation and Guidelines in Medically Assisted Reproduction. Reprod BioMed Online. 2009;18(2):15–8. https://doi.org/10.1016/S1472-6483(10)60443-9 .

Präg P, Mills MC. Cultural determinants influence assisted reproduction usage in europe more than economic and demographic factors. Hum Reprod. 2017;32(11):2305–14. https://doi.org/10.1093/humrep/dex298 .

Qiao J, Feng HL. Assisted reproductive technology in China: compliance and non-compliance. Transl Pediatr. 2014;3(2):91–7. https://doi.org/10.3978/j.issn.2224-4336.2014.01.06 .

Qinchang G, Chengliang L, Debin D. Globalization of science and international scientific collaboration: A network perspective. Geoforum. 2019;105:1–12. https://doi.org/10.1016/j.geoforum.2019.06.017 .

Sayers E (2021) Entrez Programming Utilities Help - A General Introduction to the E-utilities. Maryland: Bethesda. Retrieved from https://www.ncbi.nlm.nih.gov/books/NBK25501/

White K (2019) Publications Output: U.S. Trends and International Comparisons. Science & Engineering Indicators 2020. Arlington: National Science Foundation (NSB). Retrieved from http://files.eric.ed.gov/fulltext/ED615534.pdf

Download references

Acknowledgements

We would like to express our sincere gratitude for the financial support provided by both the Margarita Salas Fellowship from the Autonomous University of Madrid and by the Centre de recherche en éthique (CRÉ) at the University of Montreal. Both have provided invaluable financial assistance, enabling us to conduct our research.

Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature.

Author information

Authors and affiliations.

Department of Development Economics, Autonomous University of Madrid, Madrid, Spain

University of Montreal, Montreal, Canada

Ido Alon & Vardit Ravitsky

Harvard Medical School, Boston, USA

Vardit Ravitsky

SIRIS Lab, Research Division of SIRIS Academic, Barcelona, Spain

Francesco Alessandro Massucci

Department of Obstetrics and Gynaecology, Faculty of Medicine, School of Health Sciences, University of Thessaly, Larissa, Greece

Theofano Bounartzi

Mines Paris - PSL, Paris, France

Zacharie Chebance

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Ido Alon .

Ethics declarations

Ethics approval.

This research did not involve human participants or animals and did not require any informed consent of any type.

Conflicts of interests/Competing interests

We declare no conflict of interest or competing interest of any type.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix 3 Topic modeling output

Topic 1—Bioethics (green): ('0.027*"ethical" + 0.015*"genetic" + 0.014*"moral" + 0.008*"medical" + ''0.008*"legal" + 0.008*"ethic" + 0.007*"scientific" + 0.005*"principle" + ''0.005*"development" + 0.005*"medicine"').

Topic 2—Cost and Outcome (Bottom Red): ('0.040*"patient" + 0.019*"transfer" + 0.018*"rate" + 0.013*"cycle" + ''0.009*"clinical" + 0.009*"procedure" + 0.008*"risk" + 0.008*"clinic" + ''0.008*"cost" + 0.008*"ethical"').

Topic 3 – Law and Policy (Blue): ('0.024*"law" + 0.019*"legal" + 0.011*"country" + 0.010*"social" + ''0.009*"policy" + 0.008*"regulation" + 0.006*"legislation" + 0.006*"medical" '' + 0.006*"work" + 0.006*" analysis "').

Topic 4 – Psychology (Yellow): ('0.029*"couple" + 0.015*"psychological" + 0.015*"infertile" + ''0.010*"counselling" + 0.010*"factor" + 0.009*"patient" + ''0.009*"participant" + 0.008*"social" + 0.008*"relationship" + ''0.008*"interview"').

Topic 5 – Family and Sociology (Purple): ('0.092*"child" + 0.049*"parent" + 0.041*"family" + 0.022*"couple" + ''0.018*"mother" + 0.013*"lesbian" + 0.013*"conceive" + 0.012*"genetic" + ''0.012*"sex" + 0.011*"relationship"').

Topic 6 – Attitudes and Knowledge (Top Red) ('0.028*"patient" + 0.020*"age" + 0.013*"couple" + 0.013*"risk" + ''0.013*"attitude" + 0.012*"knowledge" + 0.009*"medical" + 0.008*"genetic" + ''0.007*"young" + 0.007*"participant"').

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Alon, I., Chebance, Z., Massucci, F. et al. Mapping ethical, legal, & social implications (ELSI) of assisted reproductive technologies. J Assist Reprod Genet 40 , 2045–2062 (2023). https://doi.org/10.1007/s10815-023-02854-4

Download citation

Received : 28 December 2022

Accepted : 01 June 2023

Published : 29 June 2023

Issue Date : September 2023

DOI : https://doi.org/10.1007/s10815-023-02854-4

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Assisted reproductive technologies
• Ethical social and legal implications
• Topic modeling
• Geographic distribution of research
• Find a journal
• Publish with us
• Track your research