Artificial intelligence in the in vitro fertilization laboratory: a review of advancements over the last decade

The integration of artificial intelligence (AI) and deep learning algorithms into medical care has been the focus of development over the last decade, particularly in the field of assisted reproductive technologies and in vitro fertilization (IVF). With embryo morphology the cornerstone of clinical decision making for IVF, the field of IVF is highly reliant on visual assessments that can be prone to error and subjectivity and be dependent on the level of training and expertise of the observing embryologist. Implementing AI algorithms into the IVF laboratory allows for reliable, objective, and timely assessments of both clinical parameters and microscopy images. This review discusses the ever-expanding applications of AI algorithms within the IVF embryology laboratory, aiming to discuss the many advances in multiple aspects of the IVF process. We will discuss how AI will improve various processes and procedures such as assessing oocyte quality, sperm selection, fertilization assessment, embryo assessment, ploidy prediction, embryo transfer selection, cell tracking, embryo witnessing, micromanipulation, and quality management. Overall, AI provides great potential and promise to improve not only clinical outcomes but also laboratory efficiency, a key focus because IVF clinical volume continues to increase nationwide. The integration of artificial intelligence (AI) and deep learning algorithms into medical care has been the focus of development over the last decade, particularly in the field of assisted reproductive technologies and in vitro fertilization (IVF). With embryo morphology the cornerstone of clinical decision making for IVF, the field of IVF is highly reliant on visual assessments that can be prone to error and subjectivity and be dependent on the level of training and expertise of the observing embryologist. Implementing AI algorithms into the IVF laboratory allows for reliable, objective, and timely assessments of both clinical parameters and microscopy images. This review discusses the ever-expanding applications of AI algorithms within the IVF embryology laboratory, aiming to discuss the many advances in multiple aspects of the IVF process. We will discuss how AI will improve various processes and procedures such as assessing oocyte quality, sperm selection, fertilization assessment, embryo assessment, ploidy prediction, embryo transfer selection, cell tracking, embryo witnessing, micromanipulation, and quality management. Overall, AI provides great potential and promise to improve not only clinical outcomes but also laboratory efficiency, a key focus because IVF clinical volume continues to increase nationwide. In vitro fertilization (IVF) has revolutionized reproductive medicine by providing a means for treating infertility, overcoming many diagnoses with increasing efficiency and improved outcomes over the past 40 years. The IVF laboratory is a complex environment that requires precision, accuracy, and consistency to ensure optimal outcomes. The use of artificial intelligence (AI) in the IVF laboratory is gaining momentum as a promising technology to improve the quality and efficiency of IVF procedures. This systematic review aimed to discuss the current state of advancements and future applications of AI in the IVF laboratory and its effect on various aspects of IVF procedures. In this review, we aimed to investigate the current applications of AI in the IVF laboratory. We conducted a comprehensive literature search across multiple databases, such as Web of Science, ProQuest Central, PubMed Central, and MEDLINE-Academic, using a combination of relevant keywords. The keywords used for the search were intracytoplasmic sperm injection (ICSI), IVF, oocyte, sperm, embryos, assisted reproductive technology (ART), artificial intelligence, deep learning, neural networks, machine learning (ML), and infertility. We limited our search to peer-reviewed reports published in English between the years 2010 and 2023 to focus on the most recent advances in the field of AI in the IVF laboratory. The search identified 115 results. We evaluated each reference by first screening the titles and abstracts to determine their relevance to our research question. Then, we reviewed the full-text articles of the selected references to ensure that they were peer-reviewed studies focused on the use of AI in humans within the IVF laboratory. After the initial screening, we excluded 89 articles that did not meet our inclusion criteria. Finally, we conducted a thorough evaluation of the remaining 26 articles to extract relevant information for our review. Oocyte quality is a critical factor for successful embryo development and implantation. Certain morphological characteristics of metaphase II oocytes have been shown to have clinical relevance (1Maziotis E. Sfakianoudis K. Giannelou P. Grigoriadis S. Rapani A. Tsioulou P. et al.Evaluating the value of day 0 of an ICSI cycle on indicating laboratory outcome.Sci Rep. 2020; 1019325Crossref Scopus (4) Google Scholar), and a better understanding of oocyte morphology could lead to improved patient outcomes and treatment counseling. Currently, assessment of oocyte quality is performed manually by experienced embryologists, but this process is time consuming and subjective, which may lead to inconsistencies across laboratories. The use of AI and deep learning (DL) technologies to assess oocyte quality has the potential to improve the accuracy and efficiency of this process through objective assessments with outcomes-based predictive modeling. Currently, limited studies have described using convolutional neural networks (CNNs) and support vector machines (SVMs) to analyze images of oocytes to predict their developmental potential. These technologies are used mainly to identify oocytes with higher developmental potential by analyzing images of oocytes and identifying morphological features. Kanakasabapathy et al. (2Kanakasabapathy M.K. Bormann C.L. Thirumalaraju P. Banerjee R. Shafiee H. Improving the performance of deep convolutional neural networks (CNN) in embryology using synthetic machine-generated images.Hum Reprod. 2020; 35: I209Google Scholar) developed a CNN trained to predict the likelihood of fertilization based on static oocyte images. Similarly, SVMs have been used to classify oocytes based on their morphological features (3Manna C. Nanni L. Lumini A. Pappalardo S. Artificial intelligence techniques for embryo and oocyte classification.Reprod Biomed Online. 2013; 26: 42-49Abstract Full Text Full Text PDF PubMed Scopus (70) Google Scholar, 4Targosz A. Przystałka P. Wiaderkiewicz R. Mrugacz G. Semantic segmentation of human oocyte images using deep neural networks.Biomed Eng Online. 2021; 20: 40Crossref PubMed Scopus (6) Google Scholar, 5Firuzinia S. Afzali S.M. Ghasemian F. Mirroshandel S.A. A robust deep learning-based multiclass segmentation method for analyzing human metaphase II oocyte images.Comput Methods Programs Biomed. 2021; 201105946Crossref PubMed Scopus (8) Google Scholar). By providing more accurate and reliable assessments of oocyte quality, AI can help IVF clinics better understand the factors that affect oocyte quality, leading to improved protocols and treatments in the future. The use of AI and DL technologies for sperm selection for IVF with ICSI purposes holds great promise. Traditional sperm morphology analysis has limitations owing to high interobserver variability in sperm assessment (6World Health OrganizationLaboratory manual for the examination and processing of human semen.6th ed. World Health Organization, 2021https://apps.who.int/iris/handle/10665/343208Google Scholar). An ML-based analysis offers a unique solution by using expert-trained computer processing to recognize sperm with optimal morphology. Several algorithms have been trained using manually annotated sperm data sets, which contain information regarding common classes of sperm head morphologies as reported in the WHO manual (7Chang V. Heutte L. Petitjean C. Härtel S. Hitschfeld N. Automatic classification of human sperm head morphology.Comput Biol Med. 2017; 84: 205-216Crossref PubMed Scopus (32) Google Scholar, 8Thirumalaraju P. Kanakasabapathy M.K. Bormann C.L. Kandula H. Sai Pavan S.K. Yarravarapu D. et al.Human sperm morphology analysis using smartphone microscopy and deep learning.Fertil Steril. 2019; 112: e41Abstract Full Text Full Text PDF PubMed Google Scholar, 9Riordon J. McCallum C. Sinton D. Deep learning for the classification of human sperm.Comput Biol Med. 2019; 111103342Crossref PubMed Scopus (56) Google Scholar). The use of AI and DL technologies for selecting sperm for IVF and ICSI purposes has great potential to improve the accuracy and efficiency of traditional semen analysis. AI-based approaches can recognize sperm with optimal morphology, classify sperm based on motility, and identify subtle features in brightfield images that can predict the presence of features with a negative effect on embryonic development and health. Although there are still some limitations to be addressed, the development of AI-based systems may provide ultimately useful tools for clinical embryologists to improve overall ICSI success rates. For instance, sperm motility is an essential feature measured in a conventional semen analysis. The computer-assisted sperm analysis for measuring sperm motility has become standard in many clinics worldwide. Integrating AI with video analysis offers the possibility to examine differences more accurately and rapidly based on linear movement. Recent studies suggest that AI not only matches but can also exceed the performance of standard approaches in the motility classification, providing a rapid and cost-effective alternative to conventional semen analysis (10Tsai V.F. Zhuang B. Pong Y.H. Hsieh J.T. Chang H.C. Web- and artificial intelligence-based image recognition for sperm motility analysis: verification study.JMIR Med Inform. 2020; 8e20031Crossref Scopus (12) Google Scholar, 11Ottl S. Amiriparian S. Gerczuk M. Schuller B.W. motilitAI: a machine learning framework for automatic prediction of human sperm motility.iScience. 2022; 25104644Abstract Full Text Full Text PDF PubMed Scopus (3) Google Scholar). In addition, using AI-based computer vision allows for an objective assessment of subtle differences in sperm motility patterns, potentially identifying previously unclassified motility defects for more precise sperm grade classification. Most excitingly, AI in sperm selection affords the ability to detect minute differences among images, identifying differences between sperm that may be otherwise missed by human observation alone. By performing assays for sperm function using fluorescence microscopy, although simultaneously acquiring brightfield microscopic images of sperm morphology, it is possible to identify subtle features in these brightfield images that can predict the presence of features with a negative effect on embryonic development and health. A recent study used this rationale to examine sperm DNA fragmentation, a sperm quality associated with recurrent pregnancy loss (12McCallum C. Riordon J. Wang Y. et al.Deep learning-based selection of human sperm with high DNA integrity.Commun Biol. 2019; 2: 250Crossref PubMed Scopus (39) Google Scholar). McCallum et al. (12McCallum C. Riordon J. Wang Y. et al.Deep learning-based selection of human sperm with high DNA integrity.Commun Biol. 2019; 2: 250Crossref PubMed Scopus (39) Google Scholar) trained an algorithm using acridine orange-stained sperm to identify high DNA fragmentation to predict the acridine orange level based solely on a brightfield image. This algorithm demonstrated a moderate ability to identify sperm with differing levels of DNA fragmentation (12McCallum C. Riordon J. Wang Y. et al.Deep learning-based selection of human sperm with high DNA integrity.Commun Biol. 2019; 2: 250Crossref PubMed Scopus (39) Google Scholar). These early-stage algorithms show the promise of using the objective predictive power of AI models to improve sperm selection to optimize fertilization rates and embryo development. Fertilization assessment is a critical step in the IVF process that involves evaluating the number and quality of normally fertilized eggs 14–18 hours after insemination. In a normally fertilized oocyte, the embryologist should observe the presence of 2 pronuclei (PN) within the cytoplasm of the fertilized egg. These 2 pronuclei represent the genetic material from the maternal and paternal nuclei, a sign of a normal fertilization event. Furthermore, abnormally fertilized embryos such as 0PN, 1PN, or 3PN embryos can result from unusual or abnormal combinations of genetic material. Usually, abnormally fertilized oocytes are not selected for transfer or cryopreservation because they pose an increased risk for genetic abnormalities, developmental disorders, and pregnancy loss, albeit new literature may suggest these embryos still have live birth potential (13Capalbo A. Treff N. Cimadomo D. Tao X. Ferrero S. Vaiarelli A. et al.Abnormally fertilized oocytes can result in healthy live births: improved genetic technologies for preimplantation genetic testing can be used to rescue viable embryos in in vitro fertilization cycles.Fertil Steril. 2017; 108: 1007-1015.e3Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar). The fertilization report is used commonly to make informed decisions, further emphasizing the critical nature of accurate embryo classification by embryologists. To aid in identifying PN status, Dimitriadis et al. (14Dimitriadis I. Bormann C.L. Kanakasabapathy M.K. Rice S.T. Bhowmick P. Shafiee H. et al.Deep convolutional neural networks (CNN) for assessment and selection of normally fertilized human embryos.Fertil Steril. 2019; 112: e272Abstract Full Text Full Text PDF Google Scholar) developed a CNN to distinguish between normally and abnormally fertilized oocytes, which performed at a 93.1% accuracy (14Dimitriadis I. Bormann C.L. Kanakasabapathy M.K. Rice S.T. Bhowmick P. Shafiee H. et al.Deep convolutional neural networks (CNN) for assessment and selection of normally fertilized human embryos.Fertil Steril. 2019; 112: e272Abstract Full Text Full Text PDF Google Scholar), demonstrating the utility of AI to assess fertilization. This technology has the potential to become an aid for embryologists by acting as an extra pair of “eyes” to confirm manual fertilization assessments. Currently, the use of AI and DL technologies for fertilization assessment is still experimental but has the potential to improve embryo selection and pregnancy outcomes. Studies have shown that AI algorithms can accurately predict the developmental potential of embryos based on morphokinetic parameters. For example, a recent study by Coticchio et al. (15Coticchio G. Fiorentino G. Nicora G. Sciajno R. Cavalera F. Bellazzi R. et al.Cytoplasmic movements of the early human embryo: imaging and artificial intelligence to predict blastocyst development.Reprod Biomed Online. 2021; 42: 521-528Abstract Full Text Full Text PDF PubMed Scopus (12) Google Scholar) found that analysis of cytoplasmic movements using AI can predict blastocyst development. Another study by Otsuki et al. (16Otsuki J. Iwasaki T. Enatsu N. Katada Y. Furuhashi K. Shiotani M. Noninvasive embryo selection: kinetic analysis of female and male pronuclear development to predict embryo quality and potential to produce live birth.Fertil Steril. 2019; 112: 874-881Abstract Full Text Full Text PDF PubMed Scopus (11) Google Scholar) developed a formula integrating male and female PN size to predict live birth. Although there is still much research to be performed, the use of AI in assessing fertilization has the potential to provide valuable insights into the complex processes of fertilization and early embryonic development. From the laboratory perspective, this could lead to more efficient and effective embryo selection, higher pregnancy rates, and, ultimately, better outcomes for patients undergoing IVF. IVF routinely involves culturing embryos for multiple days in controlled and monitored environmental conditions within the embryology laboratory. These embryos are assessed by highly trained embryologists through visual embryo morphology grading at different stages of embryo development to validate their quality and determine their fate. However, conventional visual evaluations of embryo morphology continues to be highly subjective, with high intervariability and intravariability, leading to inconsistencies in decision making that can compromise both patient care and advancement of the field overall (17Bormann C.L. Thirumalaraju P. Kanakasabapathy M.K. Kandula H. Souter I. Dimitriadis I. et al.Consistency and objectivity of automated embryo assessments using deep neural networks.Fertil Steril. 2020; 113: 781-787.e1Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar, 18Cimadomo D. Sosa Fernandez L. Soscia D. Fabozzi G. Benini F. Cesana A. et al.Inter-centre reliability in embryo grading across several IVF clinics is limited: implications for embryo selection.Reprod Biomed Online. 2022; 44: 39-48Abstract Full Text Full Text PDF PubMed Scopus (11) Google Scholar, 19Fordham D.E. Rosentraub D. Polsky A.L. Aviram T. Wolf Y. Perl O. et al.Embryologist agreement when assessing blastocyst implantation probability: is data-driven prediction the solution to embryo assessment subjectivity?.Hum Reprod. 2022; 37: 2275-2290Crossref PubMed Scopus (5) Google Scholar, 20Payá E. Bori L. Colomer A. Meseguer M. Naranjo V. Automatic characterization of human embryos at day 4 post-insemination from time-lapse imaging using supervised contrastive learning and inductive transfer learning techniques.Comput Methods Programs Biomed. 2022; 221106895Crossref PubMed Scopus (4) Google Scholar). To address this issue, computer-assisted evaluations could help minimize variability among embryologists in embryo scoring. Fully automated assessments of embryos, where quality grades are assigned without user interventions from an image analysis standpoint, is challenging, partly owing to the complexity of embryo morphologies. However, advances in ML have enabled objective and accurate image classification in both medical and nonmedical fields without the need for manual feature engineering. Deep learning, a subset of ML, takes a different approach through representation learning: namely, the system identifies features directly and does not rely on hand-crafted or annotated features. One of the most popular DL models that has seen numerous implementations in the field of assisted reproduction is CNNs. Such systems are trained using large data sets of static embryo images, where they learn the required parameters associated with specific qualities over time (21Khosravi P. Kazemi E. Zhan Q. Malmsten J.E. Toschi M. Zisimopoulos P. et al.Deep learning enables robust assessment and selection of human blastocysts after in vitro fertilization.NPJ Digit Med. 2019; 2: 21Crossref PubMed Scopus (190) Google Scholar). These networks far exceed human performance regarding consistency in embryo scoring and decision making, reducing variability in decision making among embryologists and improving the overall quality of patient care in the field of assisted reproduction. Bormann et al. (17Bormann C.L. Thirumalaraju P. Kanakasabapathy M.K. Kandula H. Souter I. Dimitriadis I. et al.Consistency and objectivity of automated embryo assessments using deep neural networks.Fertil Steril. 2020; 113: 781-787.e1Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar) demonstrated the power of CNN-based frameworks by conducting a study evaluating the consistency of 10 embryologists in performing routine clinical tasks such as selecting embryos for biopsy, cryopreservation, or discard. They directly compared the embryologists’ performance with the CNN-based approach and demonstrated that the CNN-based framework far exceeded human performance regarding consistency in embryo scoring and decision making. This study highlighted the potential use of CNN-based frameworks for embryo analysis in improving the overall quality of patient care in the field of assisted reproduction. The use of AI and DL for embryo quality assessment and prediction has several potential benefits for the IVF laboratory. First, it could improve the accuracy and reliability of embryo quality assessment because AI and DL algorithms can detect subtle differences in morphological features that may be indiscernible to the human eye. This could lead to more objective and standardized embryo grading, which could improve the success rates of IVF. Second, AI and DL could save time for embryologists, allowing more time for other critical tasks. Embryo quality assessment and prediction is a time-consuming process that requires considerable training and expertise. Furthermore, AI and DL could reduce the time needed for these tasks, allowing embryologists to focus on other essential tasks in the IVF laboratory. Finally, AI and DL could also reduce the workload of embryologists, allowing for level-loading of tasks as the demand for IVF services increases nationwide. In addition, AI and DL could automate some of the routine tasks, such as embryo quality assessment, allowing embryologists to focus on more complex tasks that more specifically require the manual dexterity or intellectual expertise. Ploidy refers to the number of sets of chromosomes in an organism. In IVF, ploidy status is commonly determined through preimplantation genetic testing for aneuploidy (PGT-A), a screening test requiring genetic sequencing of a trophectoderm biopsy to determine whether the embryo is euploid or aneuploid. Although euploid status can aid in selecting embryos for transfer, PGT-A is not only costly but also invasive, requiring biopsy that can lower chances of successful implantation. Artificial intelligence technologies have shown promising potential to noninvasively predict the ploidy status of an embryo by analyzing images of the embryo that are captured during the IVF process. These technologies can identify patterns and features that are associated with euploid/aneuploid embryos and use this information to make predictions (22Diakiw S.M. Hall J.M.M. VerMilyea M.D. et al.Development of an artificial intelligence model for predicting the likelihood of human embryo euploidy based on blastocyst images from multiple imaging systems during IVF.Hum Reprod. 2022; 37: 1746-1759Crossref PubMed Scopus (9) Google Scholar, 23Chavez-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. 2020; 41: 585-593Abstract Full Text Full Text PDF PubMed Scopus (41) Google Scholar, 24Huang B. Tan W. Li Z. Jin L. An artificial intelligence model (euploid prediction algorithm) can predict embryo ploidy status based on time-lapse data.Reprod Biol Endocrinol. 2021; 19: 185Crossref PubMed Scopus (17) Google Scholar, 25Cimadomo D. Marconetto A. Trio S. Chiappetta V. Innocenti F. Albricci L. et al.Human blastocyst spontaneous collapse is associated with worse morphological quality and higher degeneration and aneuploidy rates: a comprehensive analysis standardized through artificial intelligence.Hum Reprod. 2022; 37: 2291-2306Crossref PubMed Scopus (3) Google Scholar, 26Yuan Z. Yuan M. Song X. Huang X. Yan W. Development of an artificial intelligence based model for predicting the euploidy of blastocysts in PGT-A treatments.Sci Rep. 2023; 13: 2322Crossref PubMed Scopus (2) Google Scholar). Multiple groups have attempted to tackle developing an AI model that can determine euploidy. Kato et al. (27Kato K. Ueno S. Berntsen J. Kragh M.F. Okimura T. Kuroda T. Does embryo categorization by existing artificial intelligence, morphokinetic or morphological embryo selection models correlate with blastocyst euploidy rates?.Reprod Biomed Online. 2023; 46: 274-281Abstract Full Text Full Text PDF PubMed Scopus (2) Google Scholar) demonstrated that the KIDScore of blastocysts on D5 and the Gardner grade of blastocysts were associated significantly with euploidy, with a higher KIDScore and Gardner grade associated with euploidy. In addition, the study found that female age, the number of embryonic frozen days, and morphokinetic characteristics were correlated negatively with euploidy, indicating that euploidy decreased as these factors increased (27Kato K. Ueno S. Berntsen J. Kragh M.F. Okimura T. Kuroda T. Does embryo categorization by existing artificial intelligence, morphokinetic or morphological embryo selection models correlate with blastocyst euploidy rates?.Reprod Biomed Online. 2023; 46: 274-281Abstract Full Text Full Text PDF PubMed Scopus (2) Google Scholar). This study reinforced the foundation that by combining clinical parameters and morphokinetic imaging data, AI can successfully predict euploidy in a noninvasive manner. Jiang et al. (28Jiang V.S. Kandula H. Thirumalaraju P. Kanakasabapathy M.K. Cherouveim P. Souter I. et al.The use of voting ensembles to improve the accuracy of deep neural networks as a non-invasive method to predict embryo ploidy status.J Assist Reprod Genet. 2023; 40: 301-308Crossref PubMed Scopus (1) Google Scholar) described the use of an AI system that combines blastocyst images with patient characteristics to improve the accuracy of predicting embryo ploidy status. The AI system used a soft voting ensemble composed of a CNN, an SVM, and a multilayer neural network to predict embryo ploidy status through combining patient characteristics, such as maternal age, AMH level, paternal sperm quality, total number of normally fertilized embryos, and static blastocyst images. This study found that combining patient characteristics with image-based algorithms significantly improved the accuracy of euploid or aneuploid embryo classification. Similarly, 2 additional studies (29Zou Y. Pan Y. Ge N. Xu Y. Gu R. Li Z. et al.Can the combination of time-lapse parameters and clinical features predict embryonic ploidy status or implantation?.Reprod Biomed Online. 2022; 45: 643-651Abstract Full Text Full Text PDF PubMed Scopus (1) Google Scholar, 30Barnes J. Brendel M. Gao V.R. Rajendran S. Kim J. Li Q. et al.A non-invasive artificial intelligence approach for the prediction of human blastocyst ploidy: a retrospective model development and validation study.Lancet Digit Health. 2023; 5: e28-e40Abstract Full Text Full Text PDF PubMed Scopus (4) Google Scholar) demonstrated an improved ability to predict the ploidy status of embryo images when incorporating patient and clinical features into their algorithms. The development of such tools may provide a cost-efficient, standardized, and noninvasive means of embryo selection, potentially reducing health care costs and minimizing the mental, physical, and emotional burden on patients. However, there are several limitations to these studies, such as the bias introduced by the data set selection, the use of images captured only by time-lapse microscopy, and the limitation of the binary classification scheme (aneuploid vs. euploid), which could misclassify mosaic embryos with high implantation potential. Further research and validation are needed to ensure the accuracy and reliability of these noninvasive approaches for predicting ploidy status of embryos. Despite these limitations, this technology can provide standardized embryo selection and prioritization in a manner that is noninvasive, cost effective, and time efficient. Currently, AI algorithms have been developed to analyze and classify images of embryos based on various parameters such as morphology, kinetics, and genetic information. These algorithms can identify potential embryos with higher implantation and pregnancy rates, resulting in better clinical outcomes. Currently, although there are no peer-reviewed, prospective, interventional clinical trials reported to date, several retrospective studies have demonstrated the feasibility and usefulness of AI models in predicting IVF success rates (31Tran D. Cooke S. Illingworth P.J. Gardner D.K. Deep learning as a predictive tool for fetal heart pregnancy following time-lapse incubation and blastocyst transfer.Hum Reprod. 2019; 34: 1011-1018Crossref PubMed Scopus (147) Google Scholar, 32Bori L. Paya E. Alegre L. Viloria T.A. Remohi J.A. Naranjo V. et al.Novel and conventional embryo parameters as input data for artificial neural networks: an artificial intelligence model applied for prediction of the implantation potential.Fertil Steril. 2020; 114: 1232-1241Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar, 33Bormann C.L. Kanakasabapathy M.K. Thirumalaraju P. Gupta R. Pooniwala R. Kandula H. et al.Performance of a deep learning based neural network in the selection of human blastocysts for implantation.Elife. 2020; 9e55301Crossref PubMed Google Scholar). In recent years, several studies have explored the use of AI in embryo selection and grading, aiming to improve the accuracy, objectivity, and reproducibility of the process. Khosravi et al. (21Khosravi P. Kazemi E. Zhan Q. Malmsten J.E. Toschi M. Zisimopoulos P. et al.Deep learning enables robust assessment and selection of human blastocysts after in vitro fertilization.NPJ Digit Med. 2019; 2: 21Crossref PubMed Scopus (190) Google Scholar) developed “STORK,” a deep neural network (DNN) platform trained to predict blastocyst quality by embryologist annotated images of good or bad quality blastocysts. In blind tests, the STORK system achieved an impressive AUC of 0.987, with a precision score of 96%, compared with the majority opinion of the embryology team. A major advantage of this model was its fully automated process, which did not require manual annotation or pro