WISE: whole-scenario embryo identification using self-supervised learning encoder in IVF.

PURPOSE: To study the effectiveness of whole-scenario embryo identification using a self-supervised learning encoder (WISE) in in vitro fertilization (IVF) on time-lapse, cross-device, and cryo-thawed scenarios. METHODS: WISE was based on the vision transformer (ViT) architecture and masked autoencoders (MAE), a self-supervised learning (SSL) method. To train WISE, we prepared three datasets including the SSL pre-training dataset, the time-lapse identification dataset, and the cross-device identification dataset. To identify whether pairs of images were from the same embryos in different scenarios in the downstream identification tasks, embryo images including time-lapse and microscope images were first pre-processed through object detection, cropping, padding, and resizing, and then fed into WISE to get predictions. RESULTS: WISE could accurately identify embryos in the three scenarios. The accuracy was 99.89% on the time-lapse identification dataset, and 83.55% on the cross-device identification dataset. Besides, we subdivided a cryo-thawed evaluation set from the cross-device test set to have a better estimation of how WISE performs in the real-world, and it reached an accuracy of 82.22%. There were approximately 10% improvements in cross-device and cryo-thawed identification tasks after the SSL method was applied. Besides, WISE demonstrated improvements in the accuracy of 9.5%, 12%, and 18% over embryologists in the three scenarios. CONCLUSION: SSL methods can improve embryo identification accuracy even when dealing with cross-device and cryo-thawed paired images. The study is the first to apply SSL in embryo identification, and the results show the promise of WISE for future application in embryo witnessing.

Innovative technology for evaluation of sperm DNA double-strand breaks diagnoses male factor infertility and prevents reproductive failures.

Neutral comet assay has been available for two decades to evaluate sperm double-strand breaks (DSBs). However, its clinical usability is limited due to its complex and time-consuming procedure, as well as the lack of a standardized scoring system. The aim of this study was to: develop a rapid diagnostic method for DSBs, Sperm DNA Fragmentation Releasing Assay (SDFR), and explore the association between DSBs and reproductive outcomes. We pioneered the use of polyacrylamide (PA) for embedding sperm chromatin and optimized the porosity of PA to be between 10 and 13%. The refined PA network allowed the trapping of DSBs, which dispersed halo on an immunological slide; in contrast, intact chromatin failed to develop a halo. A strong correlation was showed between reproducible values obtained from SDFR and neutral comet assay. SDFR were responsive to dose-/time-dependent simulated DSBs, indicating high sensitivity and specificity. Furthermore, we conducted a retrospective study of couples with embryonic aneuploidy screening, and recording DSB profiles of the male partners. Our findings revealed that DSB enabled to predict embryonic aneuploidy whereas basic semen parameters did not. In conclusion, SDFR offers a rapid and user-friendly approach for evaluating DSBs, with potential implications for predictive healthcare in reproductive medicine.

Current trends in artificial intelligence in reproductive endocrinology.

PURPOSE OF REVIEW: Artificial Intelligence, a tool that integrates computer science and machine learning to mimic human decision-making processes, is transforming the world and changing the way we live. Recently, the healthcare industry has gradually adopted artificial intelligence in many applications and obtained some degree of success. In this review, we summarize the current applications of artificial intelligence in Reproductive Endocrinology, in both laboratory and clinical settings. RECENT FINDINGS: Artificial Intelligence has been used to select the embryos with high implantation potential, proper ploidy status, to predict later embryo development, and to increase pregnancy and live birth rates. Some studies also suggested that artificial intelligence can help improve infertility diagnosis and patient management. Recently, it has been demonstrated that artificial intelligence also plays a role in effective laboratory quality control and performance. SUMMARY: In this review, we discuss various applications of artificial intelligence in different areas of reproductive medicine. We summarize the current findings with their potentials and limitations, and also discuss the future direction for research and clinical applications.

Apoptotic qPCR gene expression array analysis demonstrates proof-of-concept for rapid blastocoel fluid-conditioned media molecular prediction.

PURPOSE: Successful identification of transcriptomic biomarkers within human IVF embryos may enhance implantation prediction and provide insights not available through conventional embryo biopsy genomic analysis. We demonstrate proof-of-concept for a methodology to assess overall embryo gene expression using qPCR with blastocoel fluid-conditioned media by examining the comparative presence of apoptotic genes. METHODS: Blastocoel fluid-conditioned media were collected from 19 embryos (11 euploid) following trophectoderm biopsy of day-5 ICSI-IVF blastocysts. Media were assessed for apoptotic gene expression via qPCR. Statistical analysis of gene expression was conducted via Wilcoxon Signed-Ranks test (overall expression), multivariate ANOVA (functional gene groups), and chi-square test of independence (gene level). RESULTS: A significantly higher overall apoptotic gene expression within euploid versus aneuploid embryos (p = 0.001) was observed. There was significantly (p = 0.045) higher expression of pro-apoptotic genes between implanted and not implanted embryos. Pro- vs. anti-apoptotic gene expression from all euploid embryos approached significance (p = 0.053). The ploidy status-based claim is further substantiated at the gene level with significantly higher expression of BBC3 (p = 0.012) and BCL2L13 (p = 0.003) in euploid embryos compared to aneuploid embryos. CONCLUSIONS: In this preliminary study, we demonstrate that (1) qualitative analysis of blastocoel fluid-conditioned media gene expression is possible, (2) global trends of expression are potentially related to clinical outcomes, and (3) gene-level expression trends exist and may be another viable metric for comparative expression between samples. The presence of statistical significance within analyses conducted with this sample size warrants a larger investigation of blastocoel fluid-conditioned media as an additional beneficial predictive tool for future IVF cases.

Protecting life in a time of war.

Despite centuries of lessons from history, war endures. Across Earth, during nearly every year from the beginning of the twentieth century to present day, over 30 wars have been fought resulting in 187 million casualties, excluding the most recent conflict, which is the impetus for this essay (Timeline of 20th and 21st century wars). We are, sadly, a war-mongering people. The word "war" word infiltrates our vernacular, e.g., the war on poverty, on drugs, on cancer, on COVID, and, apropos, on terror. How did rational approaches to disagreement and conflict evade the world's progress? Reproductive physicians and scientists are dedicated to safeguard lives and build families. Violence is antithetical to our mission as professionals, and moral integrity as humans. We are deeply concerned for, and stand in unity with, our Ukrainian colleagues-the embryologists, scientists, OBGYN and REI physicians, infertility patients, and all people under siege. Reproductive health services for Ukrainians (as with many other war-torn regions) have collapsed. Deeply disturbing reports have emerged that cite civilian hospitals (including maternity centers) being targeted. Liquid nitrogen supplies are scarce. Pregnant mothers and gestational carriers are at emergent risk of delivering in extremely harsh conditions, cold underground bunkers and refugee queues.

Common practices among consistently high-performing in vitro fertilization programs in the United States: 10-year update.

OBJECTIVE: To evaluate similarities and differences in clinical and laboratory practices among high-performing fertility clinics. DESIGN: Cross-sectional questionnaire study of selected programs. SETTING: Academic and private fertility practices performing in vitro fertilization (IVF). PATIENT(S): Not applicable. INTERVENTION(S): A comprehensive survey was conducted of 13 IVF programs performing at least 100 cycles a year and having high cumulative singleton delivery rates for 2 years. MAIN OUTCOME MEASURE(S): Clinical and laboratory IVF practices. RESULT(S): Although many areas of clinical practice varied among top programs, some commonalities were observed. All programs used a combination of follicle-stimulating hormone and luteinizing hormone for IVF stimulation, intramuscular progesterone in frozen embryo transfer cycles, ultrasound-guided embryo transfers, and a required semen analysis before starting the IVF cycle. Common laboratory practices included vitrification of embryos at the blastocyst stage, air quality control with positive air pressure and high-efficiency particulate air filtration, use of incubator gas filters, working on heated microscope stages, and incubating embryos in a low-oxygen environment, most often in benchtop incubators. CONCLUSION(S): Some areas of consistency in clinical and laboratory practices were noted among high-performing IVF programs that are likely contributing to their success. High-performing programs focused on singleton deliveries. As the field of IVF is rapidly evolving, it is imperative that we share best practices in an effort to improve outcomes from all clinics for the good of our patients.

End-to-end deep learning for recognition of ploidy status using time-lapse videos.

PURPOSE: Our retrospective study is to investigate an end-to-end deep learning model in identifying ploidy status through raw time-lapse video. METHODS: By randomly dividing the dataset of time-lapse videos with known outcome of preimplantation genetic testing for aneuploidy (PGT-A), a deep learning model on raw videos was trained by the 80% dataset, and used to test the remaining 20%, by feeding time-lapse videos as input and the PGT-A prediction as output. The performance was measured by an average area under the curve (AUC) of the receiver operating characteristic curve. RESULT(S): With 690 sets of time-lapse video image, combined with PGT-A results, our deep learning model has achieved an AUC of 0.74 from the test dataset (138 videos), in discriminating between aneuploid embryos (group 1) and others (group 2, including euploid and mosaic embryos). CONCLUSION: Our model demonstrated a proof of concept and potential in recognizing the ploidy status of tested embryos. A larger scale and further optimization on the exclusion criteria would be included in our future investigation, as well as prospective approach.

Does assisted hatching affect live birth in fresh, first cycle in vitro fertilization in good and poor prognosis patients?

PURPOSE: To assess the effect of assisted hatching (AH) on live birth rate (LBR) in first cycle, fresh in vitro fertilization (IVF) in good and poor prognosis patients. METHODS: Retrospective cohort using cycles reported to the Society for Assisted Reproductive Technology Clinic Outcomes Reporting System. Live birth rate was compared in women who underwent first cycle, autologous, fresh IVF cycles with (n = 48,858) and without (n = 103,413) AH from 2007 to 2015. RESULTS: The propensity-weighted LBR was 39.2% with AH versus 43.9% without AH in all patients. The rate difference (RD) with AH was - 4.7% ([CI - 0.053, - 0.040], P < 0.001) with the calculated number needed to harm being 22. AH affected live birth in both good prognosis and poor prognosis patients. The propensity-weighted monozygotic twinning (MZT) rate was 2.3% in patients treated with AH as compared to 1.2% patients that did not receive AH. The RD with AH on MZT in fresh, first IVF cycles was 1.1% ([0.008, 0.014], P < 0.001). CONCLUSION: AH may affect LBR across all patients and in poor prognosis patients in fresh IVF cycles. Caution should be exercised when applying this technology. More prospective research is needed.

Trophoblast differentiation, invasion and hormone secretion in a three-dimensional in vitro implantation model with rhesus monkey embryos.

BACKGROUND: The initiation of primate embryo invasion into the endometrium and the formation of the placenta from trophoblasts, fetal mesenchyme, and vascular components are essential for the establishment of a successful pregnancy. The mechanisms which direct morphogenesis of the chorionic villi, and the interactions between trophectoderm-derived trophoblasts and the fetal mesenchyme to direct these processes during placentation are not well understood due to a dearth of systems to examine and manipulate real-time primate implantation. Here we describe an in vitro three-dimensional (3-D) model to study implantation which utilized IVF-generated rhesus monkey embryos cultured in a Matrigel explant system. METHODS: Blastocyst stage embryos were embedded in a 3-D microenvironment of a Matrigel carrier and co-cultured with a feeder layer of cells generating conditioned medium. Throughout the course of embryo co-culture embryo growth and secretions were monitored. Embedded embryos were then sectioned and stained for markers of trophoblast function and differentiation. RESULTS: Signs of implantation were observed including enlargement of the embryo mass, and invasion and proliferation of trophoblast outgrowths. Expression of chorionic gonadotropin defined by immunohistochemical staining, and secretion of chorionic gonadotropin and progesterone coincident with the appearance of trophoblast outgrowths, supported the conclusion that a trophoblast cell lineage formed from implanted embryos. Positive staining for selected markers including Ki67, MHC class I, NeuN, CD31, vonWillebrand Factor and Vimentin, suggest growth and differentiation of the embryo following embedding. CONCLUSIONS: This 3-D in vitro system will facilitate further study of primate embryo biology, with potential to provide a platform for study of genes related to implantation defects and trophoblast differentiation.