Applications of artificial intelligence in ovarian stimulation: a tool for improving efficiency and outcomes
2023; Elsevier BV; Volume: 120; Issue: 1 Linguagem: Inglês
10.1016/j.fertnstert.2023.05.148
ISSN1556-5653
AutoresEduardo Hariton, Zoran J. Pavlovic, Michael Fanton, Victoria S. Jiang,
Tópico(s)Ovarian function and disorders
ResumoBecause of the birth of the first baby after in vitro fertilization (IVF), the field of assisted reproductive technologies (ARTs) has seen significant advancements in the past 40 years. Over the last decade, the healthcare industry has increasingly adopted machine learning algorithms to improve patient care and operational efficiency. Artificial intelligence (AI) in ovarian stimulation is a burgeoning niche that is currently benefiting from increased research and investment from both the scientific and technology communities, leading to cutting-edge advancements with promise for rapid clinical integration. AI-assisted IVF is a rapidly growing area of research that can improve ovarian stimulation outcomes and efficiency by optimizing the dosage and timing of medications, streamlining the IVF process, and ultimately leading to increased standardization and better clinical outcomes. This review article aims to shed light on the latest breakthroughs in this area, discuss the role of validation and potential limitations of the technology, and examine the potential of these technologies to transform the field of assisted reproductive technologies. Integrating AI responsibly into IVF stimulation will result in higher-value clinical care with the goal of having a meaningful impact on enhancing access to more successful and efficient fertility treatments. Because of the birth of the first baby after in vitro fertilization (IVF), the field of assisted reproductive technologies (ARTs) has seen significant advancements in the past 40 years. Over the last decade, the healthcare industry has increasingly adopted machine learning algorithms to improve patient care and operational efficiency. Artificial intelligence (AI) in ovarian stimulation is a burgeoning niche that is currently benefiting from increased research and investment from both the scientific and technology communities, leading to cutting-edge advancements with promise for rapid clinical integration. AI-assisted IVF is a rapidly growing area of research that can improve ovarian stimulation outcomes and efficiency by optimizing the dosage and timing of medications, streamlining the IVF process, and ultimately leading to increased standardization and better clinical outcomes. This review article aims to shed light on the latest breakthroughs in this area, discuss the role of validation and potential limitations of the technology, and examine the potential of these technologies to transform the field of assisted reproductive technologies. Integrating AI responsibly into IVF stimulation will result in higher-value clinical care with the goal of having a meaningful impact on enhancing access to more successful and efficient fertility treatments. The field of assisted reproductive technologies (ARTs) has seen significant advancements in the last 40 years, with a growing focus on improving patient outcomes through the application of new technologies (1Niederberger C. Pellicer A. Cohen J. Gardner D.K. Palermo G.D. O'Neill C.L. et al.Forty years of IVF.Fertil Steril. 2018; 110: 185-324.e5Abstract Full Text Full Text PDF PubMed Scopus (164) Google Scholar). In recent years, the healthcare industry has increased its adoption of machine learning (ML) algorithms, leveraging the benefits of computer science advancements and the management of large datasets to improve patient care and increase operational efficiency (2Davenport T. Kalakota R. The potential for artificial intelligence in healthcare.Future Healthc J. 2019; 6: 94-98Crossref PubMed Google Scholar). Applications have ranged from predicting disease outcomes and identifying optimal treatment plans to automating administrative tasks and enhancing patient engagement (2Davenport T. Kalakota R. The potential for artificial intelligence in healthcare.Future Healthc J. 2019; 6: 94-98Crossref PubMed Google Scholar). Although still in its infancy, the field of artificial intelligence (AI) in reproductive medicine has benefited from increased research and investment from both the scientific and technology communities (3Albertini D.F. The making and managing of a niche for artificial intelligence in reproductive medicine.J Assist Reprod Genet. 2023; 40: 211-212Crossref PubMed Scopus (1) Google Scholar). The potential for AI-assisted in vitro fertilization (IVF) to improve both the outcomes and efficiency of IVF through controlled ovarian stimulation is a rapidly growing area of research (4Hariton E. Chi E.A. Chi G. Morris J.R. Braatz J. Rajpurkar P. et al.A machine learning algorithm can optimize the day of trigger to improve in vitro fertilization outcomes.Fertil Steril. 2021; 116: 1227-1235Abstract Full Text Full Text PDF PubMed Scopus (10) Google Scholar, 5Letterie 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. 2020; 114: 1026-1031Abstract Full Text Full Text PDF PubMed Scopus (23) Google Scholar, 6Fanton M. Nutting V. Solano F. Maeder-York P. Hariton E. Barash O. et al.An interpretable machine learning model for predicting the optimal day of trigger during ovarian stimulation.Fertil Steril. 2022; 118: 101-108Abstract Full Text Full Text PDF PubMed Scopus (10) Google Scholar). Artificial intelligence can aid in optimizing the dosage and timing of medications, reducing the likelihood of overstimulation or under stimulation, and ultimately streamlining the IVF process. This could result in improved outcomes for patients and enable healthcare providers to treat more patients with greater efficiency and accuracy, ultimately driving increased clinical and financial value although increasing access to care. In this article, we aim to shed light on the latest breakthroughs of AI in IVF stimulation and examine the potential of these technologies to transform the field. We will briefly review the technology, extensively review existing literature, and discuss how an AI-enabled future of controlled ovarian stimulation may come to fruition. We will also discuss the role of validation in these technologies and their potential limitations. By integrating AI into the field of reproductive medicine, we expect increased access to fertility services through higher-value clinical care, leading to more successful and efficient fertility treatments. Artificial intelligence is a term that broadly describes technologies built to mimic human intelligence for tasks such as image recognition or decision-making. It has been integrated into many fields, including speech recognition and clinical decision-making support in medicine. Machine learning is a subset of AI that uses statistical modeling to analyze data patterns and generate predictive outputs without explicit coding or instructions. When paired with big data, it can lead to powerful models that draw inferences from seemingly random or inconclusive data. Machine learning models are increasingly being developed to aid clinical decision-making in healthcare. These narrow AI systems use clinical data to train algorithms for specific tasks and can range in complexity from simple linear models such as linear or logistic regression, to more complicated nonlinear algorithms such as neural networks or gradient boosting. These algorithms rely on a complex statistical analysis of clinical inputs to generate outcome predictions. There are various types of ML algorithms that perform various fundamental tasks. Regression algorithms are trained to predict continuous values, such as eggs retrieved after stimulation, whereas classification algorithms are trained to predict discrete values, such as the likelihood of blastocyst conversion or clinical pregnancy. Segmentation models are used to label or divide specific parts of an image to identify a region of interest, such as measuring the size of a follicle from an ultrasound image. Lastly, statistical techniques such as causal inference can be used to infer causal relationships between observational data and outcomes, such as understanding how the choice of stimulation protocol affects outcomes. The potential of AI in clinical decision-making has been explored across many different fields of medicine (7Kriegeskorte N. Golan T. Neural network models and deep learning.Curr Biol. 2019; 29: R231-R236Abstract Full Text Full Text PDF PubMed Scopus (170) Google Scholar, 8Deo R.C. Machine learning in medicine.Circulation. 2015; 132: 1920-1930Crossref PubMed Scopus (1460) Google Scholar). AI has emerged as a transformative force in the field of medicine, offering healthcare professionals capabilities to efficiently organize, analyze, and extract insights from the vast amounts of complex and diverse data generated daily. By harnessing the power of AI algorithms and advanced analytics, clinicians have access to more precise and timely information, enabling them to make better-informed clinical decisions and ultimately deliver better patient care. Use cases within AI are numerous, ranging from improved diagnostic capabilities in radiology to more efficient drug discovery methods that aid pharmaceutical companies in developing novel therapeutics. When it comes to diagnostic assistance, predictive analytics, or clinical decision-making support systems, AI can analyze large datasets to identify patterns and trends to predict patient outcomes or disease progression, which can help healthcare providers to individualize treatment planning and manage resources more effectively. Additionally, AI can assist physicians in making more informed decisions by providing real-time, evidence-based recommendations that can help reduce errors, improve patient outcomes, and optimize treatment plans. Artificial intelligence tools have been developed and used in various fields, including improving cancer detection in diagnostic radiology and pathology. In mammography and digital breast tomosynthesis, the use of convolutional neural networks has improved diagnostic accuracy, with sensitivity rates approaching those of radiologists (9Geras K.J. Mann R.M. Moy L. Artificial intelligence for mammography and digital breast tomosynthesis: current concepts and future perspectives.Radiology. 2019; 293: 246-259Crossref PubMed Scopus (139) Google Scholar). Convolutional neural network imaging algorithms trained to integrate clinical information from electronic health records (EHR) have shown promising results also in breast cancer detection (10Trang N.T.H. Long K.Q. An P.L. Dang T.N. Development of an artificial intelligence-based breast cancer detection model by combining mammograms and medical health records.Diagnostics (Basel). 2023; 13: 346Crossref PubMed Scopus (2) Google Scholar). Similarly, AI systems have demonstrated high concordance rates with radiologists in the detection of lung cancer (11Kim E.Y. Kim Y.J. Choi W.J. Jeon J.S. Kim M.Y. Oh D.H. et al.Concordance rate of radiologists and a commercialized deep-learning solution for chest X-ray: real-world experience with a multicenter health screening cohort.PLOS ONE. 2022; 17e0264383Google Scholar). In pathology, AI-based deep neural networks have been trained also to read needle core biopsies of prostate cancer, achieving a high area under the curve (AUC) and concordance rates with expert pathologists, indicating the potential for AI to improve diagnostic capabilities and workloads (12Ström P. Kartasalo K. Olsson H. Solorzano L. Delahunt B. Berney D.M. et al.Artificial intelligence for diagnosis and grading of prostate cancer in biopsies: a population-based, diagnostic study.Lancet Oncol. 2020; 21: 222-232Abstract Full Text Full Text PDF PubMed Scopus (260) Google Scholar). Artificial intelligence has proven to be valuable in improving patient outcomes and reducing mortality in early sepsis detection by analyzing large datasets from various sources, such as EHR, patient monitors, and laboratory results. Artificial intelligence-based platforms such as CLEW Medical's virtual intensive care units and technology electronic dashboard-intensive care units provide real-time risk assessments and alerts, although Sepsis DART developed by researchers at the University of Michigan monitors and analyzes vital signs, laboratory results, and other clinical data to identify early signs of sepsis in hospitalized patients (13Fleuren L.M. Klausch T.L.T. Zwager C.L. Schoonmade L.J. Guo T. Roggeveen L.F. et al.Machine learning for the prediction of sepsis: a systematic review and meta-analysis of diagnostic test accuracy.Intensive Care Med. 2020; 46: 383-400Crossref PubMed Scopus (201) Google Scholar, 14Gaieski D.F. Carr B. Toolan M. Ciotti K. Kidane A. Flaada D. et al.Can an end-to-end Telesepsis solution improve the severe sepsis and septic shock management Bundle-1 metrics for sepsis patients admitted from the emergency department to the hospital?.Crit Care Explor. 2022; 4e0767Crossref Google Scholar, 15Yuan 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). In drug discovery, AI accelerates the process by analyzing vast amounts of data from various sources to identify potential drug candidates and predict their effectiveness. Atomwise's AtomNet system identifies potential cancer drug candidates by analyzing protein structures, whereas BenevolentAI mines vast amounts of biomedical data to discover new drug targets and potential treatments or novel uses of currently available drugs (16Smith D.P. Oechsle O. Rawling M.J. Savory E. Lacoste A.M.B. Richardson P.J. Expert-augmented computational drug repurposing identified baricitinib as a treatment for COVID-19.Front Pharmacol. 2021; 12709856Crossref Scopus (11) Google Scholar). Artificial intelligence is poised to play a crucial role in optimizing personalized healthcare and improving clinician workflow with the use of the power of big data analysis. Artificial intelligence-based CDSS are advanced software tools that integrate AI techniques to assist decision-makers in various domains. Computer-driven support systems are designed to analyze complex data, generate insights, and provide recommendations, thereby helping professionals make more informed, accurate, and timely decisions. Letterie and Mac. (5Letterie 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. 2020; 114: 1026-1031Abstract Full Text Full Text PDF PubMed Scopus (23) Google Scholar) described a system that employs ML algorithms to analyze patient data, including age, weight, hormone levels, and ovarian reserve, to tailor the stimulation protocols to each patient's unique needs. This study consisted of 1,853 autologous and 750 donor cycles that incorporated a total of 7,376 visits for training. An additional 556 unique cycles were used to challenge the platform and calculate the accuracy of the study's algorithm. The investigators showed that the first iteration of their algorithm provided highly accurate decisions that were in concordance with evidence-based decisions by expert teams regarding whether to continue or stop treatment, trigger, and schedule oocyte retrieval or cancel a cycle, undergo medication adjustments, or predict the number of days for patient follow-up (5Letterie 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. 2020; 114: 1026-1031Abstract Full Text Full Text PDF PubMed Scopus (23) Google Scholar). Although this study did not aim to improve on or optimize decisions, it was a great proof of concept of the power of how such platforms can provide aid in reproductive medicine by providing tools for clinicians to improve treatment delivery through forming a triad between expertise, evidence, and algorithmic data analytics (17Letterie G. Three ways of knowing: the integration of clinical expertise, evidence-based medicine, and artificial intelligence in assisted reproductive technologies.J Assist Reprod Genet. 2021; 38: 1617-1625Crossref PubMed Scopus (10) Google Scholar). By providing a proof of concept that a decision-making platform can be developed that agreed with expert decisions, such platforms can be improved on and used to decrease physician workload and improve clinic efficiency for better patient care delivery. Clinics with staffing shortages or long wait times can benefit from such tools. In the development of predictive models using AI, outcome prediction is a critical component. This is because accurate prognostication before, during, or after a treatment cycle is vital for patient counseling and treatment planning. Patients face significant challenges, including psychological, financial, and psychosocial aspects, in addition to dealing with a large amount of medical information when diagnosed with infertility. The current counseling tools available are on the basis of limited studies that have poor generalizability. However, AI has the flexibility to be trained on large patient populations, allowing it to draw conclusions and identify trends that may be hidden in traditional statistical modeling. In the realm of ART, AI algorithms have been developed to predict the likelihood of clinical pregnancy and live birth using patient demographics and clinical cycle outcomes. McLernon et al. (18McLernon D.J. Raja E.A. Toner J.P. Baker V.L. Doody K.J. Seifer D.B. et al.Predicting personalized cumulative live birth following in vitro fertilization.Fertil Steril. 2022; 117: 326-338Abstract Full Text Full Text PDF PubMed Scopus (12) Google Scholar) developed logistic regression models to estimate the chances of live birth for patients going through their first cycle and for patients with an unsuccessful first cycle who are attempting a second. These models were developed from data reported to the Society for Assisted Reproductive Technology Clinic Outcome Reporting System (SART CORS), which collects fertility treatment data and outcomes from over 90% of reported IVF cycles in the United States and can be accessed freely through a calculator on the SART CORS website. Wang et al. (19Wang C.W. Kuo C.Y. Chen C.H. Hsieh Y.H. Su E.C.Y. Predicting clinical pregnancy using clinical features and machine learning algorithms in in vitro fertilization.PLOS ONE. 2022; 17e0267554Google Scholar) used a random forest algorithm to predict clinical pregnancy with an AUC of 0.72, outperforming traditional logistic regression with an AUC of 0.67. Although the AUC only had a marginal increase, the random forests algorithm they developed was able to rank predictors of clinical pregnancy, such as ovarian stimulation protocol, to assess for variable importance and the propensity of impact on positive pregnancy outcomes (19Wang C.W. Kuo C.Y. Chen C.H. Hsieh Y.H. Su E.C.Y. Predicting clinical pregnancy using clinical features and machine learning algorithms in in vitro fertilization.PLOS ONE. 2022; 17e0267554Google Scholar). Additionally, Nelson et al. (20Nelson S.M. Fleming R. Gaudoin M. Choi B. Santo-Domingo K. Yao M. Antimüllerian hormone levels and antral follicle count as prognostic indicators in a personalized prediction model of live birth.Fertil Steril. 2015; 104: 325-332Abstract Full Text Full Text PDF PubMed Scopus (31) Google Scholar) developed three validated prediction models using boosted tree modeling with antiMullerian hormone (AMH) or antral follicle count (AFC) to link cycle characteristics to live-birth probabilities. These models used clinical cycle characteristics with either AMH, AFC, or both AMH and AFC to create three distinct predictor models to assess the prognostic impact of AMH and/or AFC on clinical pregnancy. In this study, Nelson et al. (20Nelson S.M. Fleming R. Gaudoin M. Choi B. Santo-Domingo K. Yao M. Antimüllerian hormone levels and antral follicle count as prognostic indicators in a personalized prediction model of live birth.Fertil Steril. 2015; 104: 325-332Abstract Full Text Full Text PDF PubMed Scopus (31) Google Scholar) were able to show that the AMH alone model had the highest predictor power, significantly improving model predictive performance over age alone by 76.2%. Although current models are not able to predict the likelihood of clinical pregnancy on the basis of the clinical parameters alone outside of traditional logistic regression, the models described by Wang et al. (19Wang C.W. Kuo C.Y. Chen C.H. Hsieh Y.H. Su E.C.Y. Predicting clinical pregnancy using clinical features and machine learning algorithms in in vitro fertilization.PLOS ONE. 2022; 17e0267554Google Scholar) and Nelson et al. (20Nelson S.M. Fleming R. Gaudoin M. Choi B. Santo-Domingo K. Yao M. Antimüllerian hormone levels and antral follicle count as prognostic indicators in a personalized prediction model of live birth.Fertil Steril. 2015; 104: 325-332Abstract Full Text Full Text PDF PubMed Scopus (31) Google Scholar) are important stepping stones toward objectively identifying key clinical factors that may contribute to clinical pregnancy, refining variable selection for downstream algorithm development. These models can be used to personalize treatment decisions and outcomes counseling for patients and can help providers manage patient expectations and quantify financial risk in the context of multiple-cycle planning. Personalized outcome analysis can help providers better understand suboptimal outcomes using AI-based analytic tools (21Choi B. Bosch E. Lannon B.M. Leveille M.C. Wong W.H. Leader A. et al.Personalized prediction of first-cycle in vitro fertilization success.Fertil Steril. 2013; 99: 1905-1911Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). One of the few commercially available algorithms combines these concepts elegantly by training and validating a boosted tree model for IVF outcome prediction. Originally trained and validated on 5,035 IVF cycles with 52 variables (22Banerjee P. Choi B. Shahine L.K. Jun S.H. O'Leary K. Lathi R.B. et al.Deep phenotyping to predict live birth outcomes in in vitro fertilization.Proc Natl Acad Sci U S A. 2010; 107: 13570-13575Crossref PubMed Scopus (32) Google Scholar), this algorithm was further validated using an international retrospective cohort of 13,076 cycles to show an improvement of live birth probability prediction by 35.7% compared with age alone (21Choi B. Bosch E. Lannon B.M. Leveille M.C. Wong W.H. Leader A. et al.Personalized prediction of first-cycle in vitro fertilization success.Fertil Steril. 2013; 99: 1905-1911Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Additionally, the 30 significant variables included in the final analysis were assessed using a sequential multiple additive regression tree, and classification and regression tree analysis to determine their contributory impact on the performance of the model. Interestingly, they found that four variables—the total number of embryos, rate of cleavage arrest, number of 8-cell embryos, and day 3 follicle-stimulating hormone (FSH) level—accurately predicted approximately 70% of the IVF cycle outcomes, providing more informative impacts than age, clinical diagnosis, or other clinical factors (23Jun S.H. Choi B. Shahine L. Westphal L.M. Behr B. Reijo Pera R.A. et al.Defining human embryo phenotypes by cohort-specific prognostic factors.PLOS ONE. 2008; 3e2562Crossref Scopus (30) Google Scholar). These predictive tools provide a bird's-eye view to provide personalized risk stratification to predict IVF cycle outcomes, a useful counseling tool to help manage patient expectations and treatment planning. Furthermore, predictive tools that evaluate the odds of success of various treatments, such as intrauterine insemination, IVF, or third-party, may help providers and patients select the optimal treatment on the basis of their priorities. Overall, AI can play a critical role in predicting ART outcomes to ultimately enhance patient care. Given the lack of universally adopted guidelines, ovarian stimulation treatment decisions can vary significantly depending on the doctor or clinic (24Wald K. Hariton E. Morris J.R. Chi E.A. Jaswa E.G. Cedars M.I. et al.Changing stimulation protocol on repeat conventional ovarian stimulation cycles does not lead to improved laboratory outcomes.Fertil Steril. 2021; 116: 757-765Abstract Full Text Full Text PDF PubMed Scopus (4) Google Scholar). However, the aggregation of data containing this heterogeneity has enabled the training of AI models to analyze and predict how an individual will respond to different treatments, such as gonadotropin dosing and the choice of stimulation protocol. The Consolidated Standards of Reporting Trials model was one of the first approaches for using ML for gonadotropin dosing, in which a nonlinear regression model was trained on patient age, body mass index (BMI), basal FSH, and AFC to predict the dose of FSH required to retrieve 11 oocytes (25Howles C.M. Saunders H. Alam V. England P. FSH Treatment Guidelines Clinical PanelPredictive factors and a corresponding treatment algorithm for controlled ovarian stimulation in patients treated with recombinant human follicle stimulating hormone (follitropin alfa) during assisted reproduction technology (ART) procedures. An analysis of 1378 patients.Curr Med Res Opin. 2006; 22: 907-918Crossref PubMed Scopus (63) Google Scholar). However, this model has been shown in a randomized controlled trial to result in a lower number of oocytes retrieved when compared to clinician dosing (26Olivennes F. Trew G. Borini A. Broekmans F. Arriagada P. Warne D.W. et al.Randomized, controlled, open-label, non-inferiority study of the CONSORT algorithm for individualized dosing of follitropin alfa.Reprod Biomed Online. 2015; 30: 248-257Abstract Full Text Full Text PDF PubMed Scopus (35) Google Scholar). More recently, Fanton et al. (27Fanton M. Nutting V. Rothman A. Maeder-York P. Hariton E. Barash O. et al.An interpretable machine learning model for individualized gonadotrophin starting dose selection during ovarian stimulation.Reprod Biomed Online. 2022; 45: 1152-1159Abstract Full Text Full Text PDF PubMed Scopus (2) Google Scholar) developed a nearest-neighbors machine learning model to relate the starting dose of FSH to predicted mature egg outcomes. This model creates an individualized dose-response curve using a patient's age, BMI, AMH, and AFC to identify similar patients from a database of over 18,000 cycles from three US clinics. For "dose-responsive" patients that had a clear optimal region on their dose-response curve, it was estimated that selecting the predicted optimal dose could result in an average of 1.5 more MIIs. For "dose nonresponsive" patients that had a flat curve without a clear optimal region, it was estimated that selecting a low dose could result in FSH dose savings of approximately 1,375 IUs without harming outcomes. However, this approach for optimizing starting dose did not take into account other factors that can affect outcomes, such as dose adjustments midcycle or increased risk of hyperstimulation for certain populations. Other models for selecting the starting dose of gonadotropins have assumed a linear relationship between starting dose and egg outcomes. Correa et al. (28Correa N. Cerquides J. Arcos J.L. Vassena R. Supporting first FSH dosage for ovarian stimulation with machine learning.Reprod Biomed Online. 2022; 45: 1039-1045Abstract Full Text Full Text PDF PubMed Scopus (2) Google Scholar) used 2,713 patients from five clinics to predict a linear relationship between starting dose of FSH and eggs retrieved and estimated that the model could more accurately prescribe a dose resulting in 10–15 eggs compared with clinicians. Before this study, simpler linear nomograms, built off of small patient populations (N < 1,000) within a single clinic, have been proposed which used patient age, AMH, AFC, and basal FSH levels to recommend FSH dosing (29La Marca A. Papaleo E. Grisendi V. Argento C. Giulini S. Volpe A. Development of a nomogram based on markers of ovarian reserve for the individualisation of the follicle-stimulating hormone starting dose in in vitro fertilisation cycles.BJOG. 2012; 119: 1171-1179Crossref PubMed Scopus (98) Google Scholar, 30Ebid A.H.I.M. Motaleb S.M.A. Mostafa M.I. Soliman M.M.A. Novel nomogram-based integrated gonadotropin therapy individualization in in vitro fertilization/intracytoplasmic sperm injection: a modeling approach.Clin Exp Reprod Med. 2021; 48: 163-173Crossref PubMed Scopus (6) Google Scholar, 31Li Y. Duan Y. Yuan X. Cai B. Xu Y. Yuan Y. A novel nomogram for individualized gonadotropin starting dose in GnRH antagonist protocol.Front Endocrinol. 2021; 12688654Google Scholar). However, these simple linear models and nomograms are not able to capture nonlinear dose-response relationships, and make assumptions about the number of eggs retrieved considered to be "optimal." Although AI models for gonadotropin dosing have been focused primarily on the initial dose of FSH, similar models will be likely soon applied to assist with other dosing decisions, such as dosing of luteinizing hormone or predicting how gonadotropin dosing adjustments mid-stimulation will impact outcomes. Artificial intelligence techniques have been used also to better understand the choice of the stimulation protocol. A recently published article used causal inference on approximately 20,000 cycles reported to the SART CORS database between 2014–2020 to show that, for poor responders, the antagonist protocol resulted in similar outcomes to the flare protocol (32Murillo F, Fanton M, Baker VL, Loewke K. Causal inference indicates that poor responders have similar outcomes with Antagonist
Referência(s)