Predictive modeling in reproductive medicine: Where will the future of artificial intelligence research take us?

Beyond black-box models: explainable AI for embryo ploidy prediction and patient-centric consultation

PURPOSE

To determine if an explainable artificial intelligence (XAI) model enhances the accuracy and transparency of predicting embryo ploidy status based on embryonic characteristics and clinical data.

METHODS

This retrospective study utilized a dataset of 1908 blastocyst embryos. The dataset includes ploidy status, morphokinetic features, morphology grades, and 11 clinical variables. Six machine learning (ML) models including Random Forest (RF), Linear Discriminant Analysis (LDA), Logistic Regression (LR), Support Vector Machine (SVM), AdaBoost (ADA), and Light Gradient-Boosting Machine (LGBM) were trained to predict ploidy status probabilities across three distinct datasets: high-grade embryos (HGE, n = 1107), low-grade embryos (LGE, n = 364), and all-grade embryos (AGE, n = 1471). The model's performance was interpreted using XAI, including SHapley Additive exPlanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME) techniques.

RESULTS

The mean maternal age was 38.5 ± 3.85 years. The Random Forest (RF) model exhibited superior performance compared to the other five ML models, achieving an accuracy of 0.749 and an AUC of 0.808 for AGE. In the external test set, the RF model achieved an accuracy of 0.714 and an AUC of 0.750 (95% CI, 0.702-0.796). SHAP's feature impact analysis highlighted that maternal age, paternal age, time to blastocyst (tB), and day 5 morphology grade significantly impacted the predictive model. In addition, LIME offered specific case-ploidy prediction probabilities, revealing the model's assigned values for each variable within a finite range.

CONCLUSION

The model highlights the potential of using XAI algorithms to enhance ploidy prediction, optimize embryo selection as patient-centric consultation, and provides reliability and transparent insights into the decision-making process.

Use of artificial intelligence embryo selection based on static images to predict first-trimester pregnancy loss

RESEARCH QUESTION

Can an artificial intelligence embryo selection assistant predict the incidence of first-trimester spontaneous abortion using static images of IVF embryos?

DESIGN

In a blind, retrospective study, a cohort of 172 blastocysts from IVF cases with single embryo transfer and a positive biochemical pregnancy test was ranked retrospectively by the artificial intelligence morphometric algorithm ERICA. Making use of static embryo images from a light microscope, each blastocyst was assigned to one of four possible groups (optimal, good, fair or poor), and linear regression was used to correlate the results with the presence or absence of a normal fetal heart beat as an indicator of ongoing pregnancy or spontaneous abortion, respectively. Additional analyses included modelling for recipient age and chromosomal status established by preimplantation genetic testing for aneuploidy (PGT-A).

RESULTS

Embryos classified as optimal/good had a lower incidence of spontaneous abortion (16.1%) compared with embryos classified as fair/poor (25%; OR = 0.46, P = 0.005). The incidence of spontaneous abortion in chromosomally normal embryos (determined by PGT-A) was 13.3% for optimal/good embryos and 20.0% for fair/poor embryos, although the difference was not significant (P = 0.531). There was a significant association between embryo rank and recipient age (P = 0.018), in that the incidence of spontaneous abortion was unexpectedly lower in older recipients (21.3% for age ≤35 years, 17.9% for age 36-38 years, 16.4% for age ≥39 years; OR = 0.354, P = 0.0181). Overall, these results support correlation between risk of spontaneous abortion and embryo rank as determined by artificial intelligence; classification accuracy was calculated to be 67.4%.

CONCLUSIONS

This preliminary study suggests that artificial intelligence (ERICA), which was designed as a ranking system to assist with embryo transfer decisions and ploidy prediction, may also be useful to provide information for couples on the risk of spontaneous abortion. Future work will include a larger sample size and karyotyping of miscarried pregnancy tissue.

Are urologists underrepresented on fertility clinic websites? A web-based analysis

Introduction

Infertile couples frequently utilize the Internet to find various reproductive clinics and research their alternatives. Patients are increasingly using self-referral because of online information on health-care providers. The objective is to compare the image of infertility specialists to other team members on the websites of reproductive clinics.

Methods

Information was gathered during November and December 2022 from two publicly accessible online registries which include the Human Fertilization and Embryology Authority located in the United Kingdom and the Society for Assisted Reproductive Technology located in the United States. We looked over every website that was accessible, paying close attention to how each team member was portrayed online.

Results

We examined a total of 447 clinic websites. Only 8% of the profiles of male infertility doctors were included. Contrarily, most websites (96%), which specialize in reproductive endocrinology and infertility, feature the profiles of female infertility experts. Male infertility professionals also had significantly lower representation than other clinic employees, such as nurses (55.7%, P < 0.0001), directors of embryology laboratories (46.5%, P < 0.0001), office personnel (39.6%, P < 0.0001), and embryology specialists (29.7%, P < 0.0001).

Conclusion

Although male factor infertility explains the existence of over half of all cases of infertility, urologists who specialize in male infertility are glaringly understated on websites for fertility clinics. By improving this issue, fertility clinics can draw in more patients by making all members of the care team more visible.

A brief history of artificial intelligence embryo selection: from black-box to glass-box

With the exponential growth of computing power and accumulation of embryo image data in recent years, artificial intelligence (AI) is starting to be utilized in embryo selection in IVF. Amongst different AI technologies, machine learning (ML) has the potential to reduce operator-related subjectivity in embryo selection while saving labor time on this task. However, as modern deep learning (DL) techniques, a subcategory of ML, are increasingly used, its integrated black-box attracts growing concern owing to the well-recognized issues regarding lack of interpretability. Currently, there is a lack of randomized controlled trials to confirm the effectiveness of such black-box models. Recently, emerging evidence has shown underperformance of black-box models compared to the more interpretable traditional ML models in embryo selection. Meanwhile, glass-box AI, such as interpretable ML, is being increasingly promoted across a wide range of fields and is supported by its ethical advantages and technical feasibility. In this review, we propose a novel classification system for traditional and AI-driven systems from an embryology standpoint, defining different morphology-based selection approaches with an emphasis on subjectivity, explainability, and interpretability.

Artificial Intelligence in Andrology: From Semen Analysis to Image Diagnostics

Artificial intelligence (AI) in medicine has gained a lot of momentum in the last decades and has been applied to various fields of medicine. Advances in computer science, medical informatics, robotics, and the need for personalized medicine have facilitated the role of AI in modern healthcare. Similarly, as in other fields, AI applications, such as machine learning, artificial neural networks, and deep learning, have shown great potential in andrology and reproductive medicine. AI-based tools are poised to become valuable assets with abilities to support and aid in diagnosing and treating male infertility, and in improving the accuracy of patient care. These automated, AI-based predictions may offer consistency and efficiency in terms of time and cost in infertility research and clinical management. In andrology and reproductive medicine, AI has been used for objective sperm, oocyte, and embryo selection, prediction of surgical outcomes, cost-effective assessment, development of robotic surgery, and clinical decision-making systems. In the future, better integration and implementation of AI into medicine will undoubtedly lead to pioneering evidence-based breakthroughs and the reshaping of andrology and reproductive medicine.

An ultrasound-based deep learning radiomic model combined with clinical data to predict clinical pregnancy after frozen embryo transfer: a pilot cohort study

RESEARCH QUESTION

Can a multi-modal fusion model based on ultrasound-based deep learning radiomics combined with clinical parameters provide personalized evaluation of endometrial receptivity and predict the occurrence of clinical pregnancy after frozen embryo transfer (FET)?

DESIGN

Prospective cohort study of women (n = 326) who underwent FET between August 2019 and December 2021. Input quantitative variables and input image data for radiomic feature extraction were collected to establish a multi-modal fusion prediction model. An additional independent dataset of 453 ultrasound endometrial images was used to establish the segmentation model to determine the endometrial region on ultrasound images for analysis. The performance of different algorithms and different input data for prediction of FET outcome were compared.

RESULTS

A total of 240 patients with complete data were included in the final cohort. The proposed multi-modal fusion model performed significantly better than the use of either image or quantitative variables alone to predict the occurrence of clinical pregnancy after FET (P ≤ 0.034). Its area under the curve, accuracy, sensitivity, specificity, positive predictive value and negative predictive value of the proposed model were 0.825, 72.5%, 96.2%, 58.3%, 72.3% and 89.5%, respectively. The Dice coefficient of the multi-task endometrial ultrasound segmentation model was 0.89. Use of endometrial segmentation features significantly improved the prediction performance of the model (P = 0.041).

CONCLUSIONS

The multi-modal fusion model based on ultrasound-based deep learning radiomics combined with clinical quantitative variables offers a favourable and rapid non-invasive approach for personalized prediction of FET outcome.

The Renaissance of Male Infertility Management in the Golden Age of Andrology

Infertility affects nearly 186 million people worldwide and the male partner is the cause in about half of the cases. Meta-regression data indicate an unexplained decline in sperm concentration and total sperm count over the last four decades, with an increasing prevalence of male infertility. This suggests an urgent need to implement further basic and clinical research in Andrology. Andrology developed as a branch of urology, gynecology, endocrinology, and, dermatology. The first scientific journal devoted to andrological sciences was founded in 1969. Since then, despite great advancements, andrology has encountered several obstacles in its growth. In fact, for cultural reasons, the male partner has often been neglected in the diagnostic and therapeutic workup of the infertile couple. Furthermore, the development of assisted reproductive techniques (ART) has driven a strong impression that this biotechnology can overcome all forms of infertility, with a common belief that having a spermatozoon from a male partner (a sort of sperm donor) is all that is needed to achieve pregnancy. However, clinical practice has shown that the quality of the male gamete is important for a successful ART outcome. Furthermore, the safety of ART has been questioned because of the high prevalence of comorbidities in the offspring of ART conceptions compared to spontaneous conceptions. These issues have paved the way for more research and a greater understanding of the mechanisms of spermatogenesis and male infertility. Consequently, numerous discoveries have been made in the field of andrology, ranging from genetics to several "omics" technologies, oxidative stress and sperm DNA fragmentation, the sixth edition of the WHO manual, artificial intelligence, management of azoospermia, fertility in cancers survivors, artificial testis, 3D printing, gene engineering, stem cells therapy for spermatogenesis, and reconstructive microsurgery and seminal microbiome. Nevertheless, as many cases of male infertility remain idiopathic, further studies are required to improve the clinical management of infertile males. A multidisciplinary strategy involving both clinicians and scientists in basic, translational, and clinical research is the core principle that will allow andrology to overcome its limits and reach further goals. This state-of-the-art article aims to present a historical review of andrology, and, particularly, male infertility, from its "Middle Ages" to its "Renaissance", a golden age of andrology.

Automation in ART: Paving the Way for the Future of Infertility Treatment

In vitro fertilisation (IVF) is estimated to account for the birth of more than nine million babies worldwide, perhaps making it one of the most intriguing as well as commoditised and industrialised modern medical interventions. Nevertheless, most IVF procedures are currently limited by accessibility, affordability and most importantly multistep, labour-intensive, technically challenging processes undertaken by skilled professionals. Therefore, in order to sustain the exponential demand for IVF on one hand, and streamline existing processes on the other, innovation is essential. This may not only effectively manage clinical time but also reduce cost, thereby increasing accessibility, affordability and efficiency. Recent years have seen a diverse range of technologies, some integrated with artificial intelligence, throughout the IVF pathway, which promise personalisation and, at least, partial automation in the not-so-distant future. This review aims to summarise the rapidly evolving state of these innovations in automation, with or without the integration of artificial intelligence, encompassing the patient treatment pathway, gamete/embryo selection, endometrial evaluation and cryopreservation of gametes/embryos. Additionally, it shall highlight the resulting prospective change in the role of IVF professionals and challenges of implementation of some of these technologies, thereby aiming to motivate continued research in this field.

Prediction of oocyte maturation rate in the GnRH antagonist flexible IVF protocol using a novel machine learning algorithm - A retrospective study

Oocyte maturation is affected by various patient and cycle parameters and has a key effect on treatment outcome. A prediction model for oocyte maturation rate formulated by using machine learning and neural network algorithms has not yet been described. A retrospective cohort study that included all women aged ≤ 38 years who underwent their first IVF treatment using a flexible GnRH antagonist protocol in a single tertiary hospital between 2010 and 2015. 462 patients met the inclusion criteria. Median maturation rate was approximately 80%. Baseline characteristics and treatment parameters of cycles with high oocyte maturation rate (≥80%, n = 236) were compared to cycles with low oocyte maturation rate (<80%, n = 226). We used an XGBoost algorithm that fits the training data using decision trees and rates factors according to their influence on the prediction. For the machine training phase, 80% of the cohort was randomly selected, while rest of the samples were used to evaluate our model's accuracy. We demonstrated an accuracy rate of 75% in predicting high oocyte maturation rate in GnRH antagonist cycles. Our model showed an operating characteristic curve with AUC of 0.78 (95% CI 0.73-0.82). The most predictive parameters were peak estradiol level on trigger day, estradiol level on antagonist initiation day, average dose of gonadotropins per day and progesterone level on trigger day. A state-of-the-art machine learning algorithm presented promising ability to predict oocyte maturation rate in the first GnRH antagonist flexible protocol using simple parameters before final trigger for ovulation. A prospective study to evaluate this model is needed.

The Impact of Digital Health on Smoking Cessation

BACKGROUND

Smartphones have become useful tools for medicine, with the use of specific apps making it possible to bring health care closer to inaccessible areas, continuously monitor a patient's pathology at any time and place, promote healthy habits, and ultimately improve patients' quality of life and the efficiency of the health care system. Since 2020, the use of smartphones has reached unprecedented levels. There are more than 350,000 health apps, according to a 2021 IQVIA Institute report, that address, among other things, the management of patient appointments; communication among different services or professionals; the promotion of lifestyle changes related to adopting healthy habits; and the monitoring of different pathologies and chronic conditions, including smoking cessation. The number of mobile apps for quitting smoking is high. As early as 2017, a total of 177 unique smoking cessation-relevant apps were identified in the iPhone App Store, 139 were identified in Google Play, 70 were identified in the BlackBerry app store, and 55 were identified in the Windows Phone Store, but very few have adequate scientific support. It seems clear that efforts are needed to assess the quality of these apps, as well as their effectiveness in different population groups, to have tools that offer added value to standard practices.

OBJECTIVE

This viewpoint aims to highlight the benefits of mobile health (mHealth) and its potential as an adjuvant tool in health care.

METHODS

A review of literature and other data sources was performed in order to show the current status of mobile apps that can offer support for smoking cessation. For this purpose, the PubMed, Embase, and Cochrane databases were explored between May and November 2022.

RESULTS

In terms of smoking cessation, mHealth has become a powerful coadjuvant tool that allows health workers to perform exhaustive follow-ups for the process of quitting tobacco and provide support anytime and anywhere. mHealth tools are effective for different groups of smokers (eg, pregnant women, patients with chronic obstructive pulmonary disease, patients with mental illness, and the general population) and are cost-effective, generating savings for the health system. However, there are some patient characteristics that can predict the success of using mobile apps in the smoking cessation process, such as the lower age of patients, dependence on tobacco, the number of quit attempts, and the previous use of mobile apps, among others. Therefore, it is preferable to offer these tools to patients with a higher probability of quitting tobacco.

CONCLUSIONS

mHealth is a promising tool for helping smokers in the smoking cessation process. There is a need for well-designed clinical studies and economic evaluations to jointly assess the effectiveness of new interventions in different population groups, as well as their impact on health care resources.

Development and validation of deep learning based embryo selection across multiple days of transfer

This work describes the development and validation of a fully automated deep learning model, iDAScore v2.0, for the evaluation of human embryos incubated for 2, 3, and 5 or more days. We trained and evaluated the model on an extensive and diverse dataset including 181,428 embryos from 22 IVF clinics across the world. To discriminate the transferred embryos with known outcome, we show areas under the receiver operating curve ranging from 0.621 to 0.707 depending on the day of transfer. Predictive performance increased over time and showed a strong correlation with morphokinetic parameters. The model's performance is equivalent to the KIDScore D3 model on day 3 embryos while it significantly surpasses the performance of KIDScore D5 v3 on day 5+ embryos. This model provides an analysis of time-lapse sequences without the need for user input, and provides a reliable method for ranking embryos for their likelihood of implantation, at both cleavage and blastocyst stages. This greatly improves embryo grading consistency and saves time compared to traditional embryo evaluation methods.

How to Improve Non-Invasive Diagnosis of Endometriosis with Advanced Statistical Methods

Background and Objectives: Endometriosis is one of the most common gynecological disorders in women of reproductive age. Causing pelvic pain and infertility, it is considered one of the most serious health problems, being responsible for work absences or productivity loss. Its diagnosis is often delayed because of the need for an invasive laparoscopic approach. Despite years of studies, no single marker for endometriosis has been discovered. The aim of this research was to find an algorithm based on symptoms and laboratory tests that could diagnose endometriosis in a non-invasive way. Materials and Methods: The research group consisted of 101 women hospitalized for diagnostic laparoscopy, among which 71 had confirmed endometriosis. Data on reproductive history were collected in detail. CA125 (cancer antigen-125) level and VEGF1(vascular endothelial growth factor 1) were tested in blood samples. Among the used statistical methods, the LASSO regression-a new important statistical tool eliminating the least useful features-was the only method to have significant results. Results: Out of 19 features based on results of LASSO, 7 variables were chosen: body mass index, age of menarche, cycle length, painful periods, information about using contraception, CA125, and VEGF1. After multivariate logistic regression with a backward strategy, the three most significant features were evaluated. The strongest impact on endometriosis prediction had information about painful periods, CA125 over 15 u/mL, and the lowest BMI, with a sensitivity of 0.8800 and a specificity of 0.8000, respectively. Conclusions: Advanced statistical methods are crucial when creating non-invasive tests for endometriosis. An algorithm based on three easy features, including painful menses, BMI level, and CA125 concentration could have an important place in the non-invasive diagnosis of endometriosis. If confirmed in a prospective study, implementing such an algorithm in populations with a high risk of endometriosis will allow us to cover patients suspected of endometriosis with proper treatment.

New frontiers in embryo selection

Human infertility is a major global public health issue estimated to affect one out of six couples, while the number of assisted reproduction cycles grows impressively year over year. Efforts to alleviate infertility using advanced technology are gaining traction rapidly as infertility has an enormous impact on couples and the potential to destabilize entire societies if replacement birthrates are not achieved. Artificial intelligence (AI) technologies, leveraged by the highly advanced assisted reproductive technology (ART) industry, are a promising addition to the armamentarium of tools available to combat global infertility. This review provides a background for current methodologies in embryo selection, which is a manual, time-consuming, and poorly reproducible task. AI has the potential to improve this process (among many others) in both the clinician's office and the IVF laboratory. Embryo selection is evolving through digital methodologies into an automated procedure, with superior reliability and reproducibility, that is likely to result in higher pregnancy rates for patients. There is an emerging body of data demonstrating the utility of AI applications in multiple areas in the IVF laboratory. AI platforms have been developed to evaluate individual embryologist performance; to provide quality assurance for culture systems; to correlate embryologist's assessments and AI systems; to predict embryo ploidy, implantation, fetal heartbeat, and live birth outcome; and to replace the current "analogue" system of embryo selection with a digital paradigm. AI capability will distinguish high performing, high profit margin, low-cost, and highly successful IVF clinic business models. We think it will become the standard, "new normal" in IVF labs, as rapidly and thoroughly as vitrification, blastocyst culture, and intracytoplasmic sperm injection replaced their predecessor technologies. At the time of this review, the AI technology to automate embryo evaluation and selection has robustly matured, and therefore, it is the main focus of this review.

Could artificial intelligence revolutionize the development of nanovectors for gene therapy and mRNA vaccines?

Gene therapy enables the introduction of nucleic acids like DNA and RNA into host cells, and is expected to revolutionize the treatment of a wide range of diseases. This growth has been further accelerated by the discovery of CRISPR/Cas technology, which allows accurate genomic editing in a broad range of cells and organisms in vitro and in vivo. Despite many advances in gene delivery and the development of various viral and non-viral gene delivery vectors, the lack of highly efficient non-viral systems with low cellular toxicity remains a challenge. The application of cutting-edge technologies such as artificial intelligence (AI) has great potential to find new paradigms to solve this issue. Herein, we review AI and its major subfields including machine learning (ML), neural networks (NNs), expert systems, deep learning (DL), computer vision and robotics. We discuss the potential of AI-based models and algorithms in the design of targeted gene delivery vehicles capable of crossing extracellular and intracellular barriers by viral mimicry strategies. We finally discuss the role of AI in improving the function of CRISPR/Cas systems, developing novel nanobots, and mRNA vaccine carriers.

SART CORS IVF registry: looking to the past to shape future perspectives

PURPOSE

The SART CORS database is an informative source of IVF clinic-specific linked data that provides cumulative live birth rates from medically assisted reproduction in the United States (US). These data are used to develop best practice guidelines, for research, quality assurance, and post-market surveillance of assisted reproductive technologies. Here, we sought to investigate the key areas of current research focus (higher-order categories), discover gaps or underserved areas of ART research, and examine the potential application and impact of newer ART adjuvants, future data collection, and analysis needs.

METHODS

We conducted a systematic review (PRISMA guidelines) to quantify unique output metrics of the SART CORS database. Included were SART member reporting clinics: full-length publications from 2004 to 2021 and conference abstracts from 2015 to 2021, the two key timepoints when the SART CORS database underwent transformative shifts in data collection.

RESULTS

We found 206 abstracts presented from 2015 to 2021, 189 full-length peer-reviewed publications since 2004, with 654 unique authors listed on these publications. A total of 19 publications have been highly impactful, garnering over 100 citations at the time of writing. Several higher-order categories, such as endometriosis and tubal infertility, have few publications. The conversion of conference abstracts to full-length papers ranged from 15 to 35% from 2015 to 2021.

CONCLUSIONS

A substantial body of literature has been generated by analyzing the SART CORS database. Full-length publications have increased year over year. Some topic areas, such as endometriosis and tubal infertility, may be underrepresented. Conversion of conference abstracts to full-length publications has been low, indicating that more organizational support may be needed to ensure that research is methodologically sound and researchers supported to reach full publication status.

Should we adopt a prognosis-based approach to unexplained infertility?

The treatment of unexplained infertility is a contentious topic that continues to attract a great deal of interest amongst clinicians, patients and policy makers. The inability to identify an underlying pathology makes it difficult to devise effective treatments for this condition. Couples with unexplained infertility can conceive on their own and any proposed intervention needs to offer a better chance of having a baby. Over the years, several prognostic and prediction models based on routinely collected clinical data have been developed, but these are not widely used by clinicians and patients. In this opinion paper, we propose a prognosis-based approach such that a decision to access treatment is based on the estimated chances of natural and treatment-related conception, which, in the same couple, can change over time. This approach avoids treating all couples as a homogeneous group and minimizes unnecessary treatment whilst ensuring access to those who need it early.

A Federated Blockchain Approach for Fertility Preservation and Assisted Reproduction in Smart Cities

Modern life is making people infertile. Giving birth later in life is wreaking havoc on our fertility and threatening human survival. Smart cities intend to optimize the quality of life of their citizens by utilizing technology for smarter living. This research first identifies the requirements and business opportunities of using advanced technology for smarter fertility preservation and assisted reproduction in smart cities. A federated blockchain approach is proposed for the alliance of integrated commercial egg banks (ICEBs). In particular, we designed a membership fee rebate (MFR) mechanism that offers incentives for blockchain creations in the egg banking alliance. We formulated the MFR problem into a leader–followers Stackelberg game whose objectives are (1) to maximize the benefits of forming the alliance (the leader) and (2) to maximize the benefits in each ICEB (the follower). We developed an iterative scheme that utilizes mathematical programming techniques to solve the two-level, Stackelberg game problem. With a given set of parameters of the alliance and membership fee function, and the average number of blocks generated for an oocyte, the iterative scheme achieves the optimal solution for the MFR rate per block created. A numerical example demonstrates the feasibility and applicability of the proposed iterative scheme. Numerical results show that it achieves good solutions in adding a small to medium-sized new ICEB to the existing alliance. The proposed federated approach lays the foundation for developing a blockchain-based egg banking platform.

Vanquishing multiple pregnancy in in vitro fertilization in the United States-a 25-year endeavor

The practice of in vitro fertilization has changed tremendously since the birth of the first in vitro fertilization infant in 1978. With the success of early in vitro fertilization programs in the United States, there was a substantial rise in twin births nationwide. In the mid-1990s, more than 30% of in vitro fertilization cycles resulted in twin or higher-order multifetal pregnancies. Since that time, we not only have witnessed improvements in laboratory and treatment efficacy but also have seen a dramatic impact on pregnancy outcomes, specifically regarding twin pregnancies. Because the field evolved and the risks of multifetal pregnancies became more salient, in 2019, the rate of twin pregnancies had dropped to <7% of cycles. This improvement was largely because of technical advancements and revised professional guidance: culturing embryos longer before transfer, improved freezing technology, embryo preimplantation genetic testing, and revised professional guidance regarding the number of embryos to transfer. These developments have led to single-embryo transfer becoming the standard of care in most scenarios. We used national in vitro fertilization surveillance data of all autologous in vitro fertilization cycles from 1996 to 2019 to illustrate trends in the following improved outcomes: autologous embryo transfer cycles involving blastocyst-stage embryos, vitrified embryos, preimplantation genetic testing cycles, total number of embryos being transferred per cycle, and single-embryo transfer usage over time. Among deliveries from autologous embryo transfers, we highlighted trends in singleton births over time and proportion of deliveries involving twins, triplets, quadruplets, or greater. The notable progress in reducing the rate of multifetal pregnancies with in vitro fertilization was largely attributed to a series of technical and clinical actions, culminating in an 80% reduction in the incidence of multiple births without a loss in overall treatment effectiveness.

Predicting Embryo Viability Based on Self-Supervised Alignment of Time-Lapse Videos

With self-supervised learning, both labeled and unlabeled data can be used for representation learning and model pretraining. This is particularly relevant when automating the selection of a patient's fertilized eggs (embryos) during a fertility treatment, in which only the embryos that were transferred to the female uterus may have labels of pregnancy. In this paper, we apply a self-supervised video alignment method known as temporal cycle-consistency (TCC) on 38176 time-lapse videos of developing embryos, of which 14550 were labeled. We show how TCC can be used to extract temporal similarities between embryo videos and use these for predicting pregnancy likelihood. Our temporal similarity method outperforms the time alignment measurement (TAM) with an area under the receiver operating characteristic (AUC) of 0.64 vs. 0.56. Compared to existing embryo evaluation models, it places in between a pure temporal and a spatio-temporal model that both require manual annotations. Furthermore, we use TCC for transfer learning in a semi-supervised fashion and show significant performance improvements compared to standard supervised learning, when only a small subset of the dataset is labeled. Specifically, two variants of transfer learning both achieve an AUC of 0.66 compared to 0.63 for supervised learning when 16% of the dataset is labeled.

Pregnancy prediction performance of an annotation-free embryo scoring system on the basis of deep learning after single vitrified-warmed blastocyst transfer: a single-center large cohort retrospective study

OBJECTIVE

To analyze the performance of an annotation-free embryo scoring system on the basis of deep learning for pregnancy prediction after single vitrified blastocyst transfer (SVBT) compared with the performance of other blastocyst grading systems dependent on annotation or morphology scores.

DESIGN

A single-center large cohort retrospective study from an independent validation test.

SETTING

Infertility clinic.

PATIENT(S)

Patients who underwent SVBT cycles (3,018 cycles, mean ± SD patient age 39.3 ± 4.0 years).

INTERVENTION(S)

None.

MAIN OUTCOME MEASURE(S)

The pregnancy prediction performances of each embryo scoring model were compared using the area under curve (AUC) for predicting the fetal heartbeat status for each maternal age group.

RESULT(S)

The AUCs of the <35 years age group (n = 389) for pregnancy prediction were 0.72 for iDAScore, 0.66 for KIDScore, and 0.64 for the Gardner criteria. The AUC of iDAScore was significantly greater than those of the other two models. For the 35-37 years age group (n = 514), the AUCs were 0.68, 0.68, and 0.65 for iDAScore, KIDScore, and the Gardner criteria, respectively, and were not significantly different. The AUCs of the 38-40 years age group (n = 796) were 0.67 for iDAScore, 0.65 for KIDScore, and 0.64 for the Gardner criteria, and there were no significant differences. The AUCs of the 41-42 years age group (n = 636) were 0.66, 0.66, and 0.63 for iDAScore, KIDScore, and the Gardner criteria, respectively, and there were no significant differences among the pregnancy prediction models. For the >42 years age group (n = 389), the AUCs were 0.76 for iDAScore, 0.75 for KIDScore, and 0.75 for the Gardner criteria, and there were no significant differences. Thus, iDAScore AUC was either the highest or equal to the highest AUC for all age groups, although a significant difference was observed only in the youngest age group.

CONCLUSION(S)

Our results showed that objective embryo assessment by a completely automatic and annotation-free model, iDAScore, performed as well as or even better than more traditional embryo assessment or annotation-dependent ranking tools. iDAScore could be an optimal pregnancy prediction model after SVBT, especially in young patients.

Embryo selection with artificial intelligence: how to evaluate and compare methods?

Embryo selection within in vitro fertilization (IVF) is the process of evaluating qualities of fertilized oocytes (embryos) and selecting the best embryo(s) available within a patient cohort for subsequent transfer or cryopreservation. In recent years, artificial intelligence (AI) has been used extensively to improve and automate the embryo ranking and selection procedure by extracting relevant information from embryo microscopy images. The AI models are evaluated based on their ability to identify the embryo(s) with the highest chance(s) of achieving a successful pregnancy. Whether such evaluations should be based on ranking performance or pregnancy prediction, however, seems to divide studies. As such, a variety of performance metrics are reported, and comparisons between studies are often made on different outcomes and data foundations. Moreover, superiority of AI methods over manual human evaluation is often claimed based on retrospective data, without any mentions of potential bias. In this paper, we provide a technical view on some of the major topics that divide how current AI models are trained, evaluated and compared. We explain and discuss the most common evaluation metrics and relate them to the two separate evaluation objectives, ranking and prediction. We also discuss when and how to compare AI models across studies and explain in detail how a selection bias is inevitable when comparing AI models against current embryo selection practice in retrospective cohort studies.

Artificial intelligence-the future is now

The pros and cons of artificial intelligence in assisted reproductive technology are presented.