Deep learning system for classification of ploidy status using time-lapse videos.

OBJECTIVE: To develop a spatiotemporal model for de prediction of euploid and aneuploid embryos using time-lapse videos from 10-115 hours after insemination (hpi). DESIGN: Retrospective study. MAIN OUTCOME MEASURES: The research used an end-to-end approach to develop an automated artificial intelligence system capable of extracting features from images and classifying them, considering spatiotemporal dependencies. A convolutional neural network extracted the most relevant features from each video frame. A bidirectional long short-term memory layer received this information and analyzed the temporal dependencies, obtaining a low-dimensional feature vector that characterized each video. A multilayer perceptron classified them into 2 groups, euploid and noneuploid. RESULTS: The model performance in accuracy fell between 0.6170 and 0.7308. A multi-input model with a gate recurrent unit module performed better than others; the precision (or positive predictive value) is 0.8205 for predicting euploidy. Sensitivity, specificity, F1-Score and accuracy are 0.6957, 0.7813, 0.7042, and 0.7308, respectively. CONCLUSIONS: This article proposes an artificial intelligence solution for prioritizing euploid embryo transfer. We can highlight the identification of a noninvasive method for chromosomal status diagnosis using a deep learning approach that analyzes raw data provided by time-lapse incubators. This method demonstrated potential automation of the evaluation process, allowing spatial and temporal information to encode.

A propensity score-based, comparative study assessing humid and dry time-lapse incubation, with single-step medium, on embryo development and clinical outcomes.

STUDY QUESTION: Does culture in a high relative humidity atmosphere improve clinical outcomes when using a time-lapse integrated incubator and single-step culture medium? SUMMARY ANSWER: Using an integrated time-lapse system and single-step culture medium, culture in a high relative humidity atmosphere increases the likelihood of embryos, especially those subjected to preimplantation genetic testing for aneuploidies, to achieve a pregnancy compared to those cultured in dry conditions. WHAT IS KNOWN ALREADY: The use of a humid atmosphere inside incubators can reduce changes in culture media osmolality, which has been reported to have a significant effect on embryo quality and morphokinetics. Studies assessing the effect of humid culture (HC) in clinical outcomes are, however, scarce and inconclusive, mostly due to a high variability in culture conditions and reduced sample size. STUDY DESIGN, SIZE, DURATION: Retrospective cohort study performed over 1627 ICSI cycles performed during 3 consecutive years in which embryo cohorts were cultured in a time-lapse incubator with three dry and three humidified chambers, and using single-step culture medium. Clinical outcomes were compared between treatments in which embryo cohorts were cultured in either humid (n = 833) or dry (n = 794) conditions. PARTICIPANTS/MATERIALS, SETTING, METHODS: The study includes autologous treatments, with (N = 492) and without (N = 372) preimplantation genetic testing for aneuploidies (PGT-A) and ovum donation treatments (N = 763), performed in three university-affiliated private IVF centres. Stimulation, oocyte pickup and fertilization were performed according to the standard procedures of the clinic. All embryo cohorts were cultured in the same model of time-lapse incubator, distributed to either a dry or humidified chamber, while the rest of the culture variables remained equal. The population was weighted by the inverse probability of treatment to control for all measured confounders. The association between HC and the main outcome was assessed by logistic regression over the weighted population. The E-value was reported as a way of considering for unmeasured confounders. Differences in embryo development and other secondary outcomes between the study groups were assessed by Pearson Chi-squared test, ANOVA test and Kaplan-Meier survival analysis. MAIN RESULTS AND THE ROLE OF CHANCE: An univariable logistic regression analysis, weighted by the inverse probability of treatment, determined that embryos cultured in humid conditions are more likely to achieve a clinical pregnancy than those cultured in dry conditions (odds ratio (OR) = 1.236 (95% CI 1.009-1.515), P = 0.041, E = 1.460). Through stratification, it was determined that said effect is dependent on the type of treatment: no improvement in clinical pregnancy was present in ovum donation or autologous treatments, but a statistically significant positive effect was present in treatments with preimplantation genetic testing (OR = 1.699 (95% CI 1.084-2.663), P = 0.021, E = 1.930). Said increase does not relate with an improvement in later outcomes. Differences were also found in variables related to embryo developmental morphokinetics. LIMITATIONS, REASONS FOR CAUTION: The retrospective nature of the study makes it susceptible to some bias linked to the characteristics of the treatments. To lessen the effect of possible biases, cases were weighted by the inverse probability of treatment prior to the evaluation of the outcome, as means to assess for measured confounders. In addition, the E-value of the weighted OR was calculated as a sensitivity analysis for unmeasured confounders. A randomized prospective study could be performed for further assessing the effect of humid conditions in clinical outcome. WIDER IMPLICATIONS OF THE FINDINGS: These results support that embryo culture under conditions of high relative humidity contributes to optimize clinical results in undisturbed culture in a time-lapse incubator with single-step medium. To our knowledge, this is the largest study on the matter and the first performing a propensity score-based analysis. STUDY FUNDING/COMPETING INTEREST(S): This work was supported by the ''Centro para el Desarrollo Tecnologico Industrial'' from the Spanish Ministry of Science, Innovation, and Universities (CDTI-20170310) and Generalitat Valenciana and European Social Fund (ACIF/2019/264). None of the authors have any competing interest to declare. TRIAL REGISTRATION NUMBER: N/A.

Automatic characterization of human embryos at day 4 post-insemination from time-lapse imaging using supervised contrastive learning and inductive transfer learning techniques.

BACKGROUND: Embryo morphology is a predictive marker for implantation success and ultimately live births. Viability evaluation and quality grading are commonly used to select the embryo with the highest implantation potential. However, the traditional method of manual embryo assessment is time-consuming and highly susceptible to inter- and intra-observer variability. Automation of this process results in more objective and accurate predictions. METHOD: In this paper, we propose a novel methodology based on deep learning to automatically evaluate the morphological appearance of human embryos from time-lapse imaging. A supervised contrastive learning framework is implemented to predict embryo viability at day 4 and day 5, and an inductive transfer approach is applied to classify embryo quality at both times. RESULTS: Results showed that both methods outperformed conventional approaches and improved state-of-the-art embryology results for an independent test set. The viability result achieved an accuracy of 0.8103 and 0.9330 and the quality results reached values of 0.7500 and 0.8001 for day 4 and day 5, respectively. Furthermore, qualitative results kept consistency with the clinical interpretation. CONCLUSIONS: The proposed methods are up to date with the artificial intelligence literature and have been proven to be promising. Furthermore, our findings represent a breakthrough in the field of embryology in that they study the possibilities of embryo selection at day 4. Moreover, the grad-CAMs findings are directly in line with embryologists' decisions. Finally, our results demonstrated excellent potential for the inclusion of the models in clinical practice.

The higher the score, the better the clinical outcome: retrospective evaluation of automatic embryo grading as a support tool for embryo selection in IVF laboratories.

STUDY QUESTION: Is the automatic embryo grading function of specific time-lapse systems clinically useful as a decision support tool for IVF laboratories? SUMMARY ANSWER: Blastocyst grading according to the automatic scoring system is directly associated with the likelihood of implantation and live birth, at least in treatments without preimplantation genetic testing for aneuploidy (PGT-A). WHAT IS KNOWN ALREADY: Several embryo selection algorithms have been described since the introduction of time-lapse technology in IVF laboratories, but no one algorithm has yet been sufficiently consolidated for universal use. Multicentric models based on automated grading systems offer promise for standardization of embryo selection. STUDY DESIGN, SIZE, DURATION: A retrospective cohort study was performed including 1678 patients who underwent IVF treatments between 2018 and 2020 and whose embryos (n = 12 468) were cultured in time-lapse systems. PARTICIPANTS/MATERIALS, SETTING, METHODS: After obtaining the required parameters (division time to 2, 3, 4 and 5 cells; time of blastocyst formation; inner cell mass quality; and trophectoderm quality), the automatic embryo score was calculated using the software included in the appropriate workstation. First, embryo score was compared with conventional morphological quality and the subsequent clinical outcomes of 1952 single blastocyst transfers. Second, we quantified the contribution of the automatic embryo score and conventional morphological grade to implantation and live birth outcome with multivariate logistic regression analysis in different patient populations. MAIN RESULTS AND THE ROLE OF CHANCE: A higher embryo score was associated with a better clinical outcome of IVF treatment. The mean of the automatic embryo score varied significantly (P < 0.001) among embryos with different morphological categories, between euploid and aneuploid embryos, between embryos resulting in positive versus negative pregnancy, between implanted and non-implanted embryos, and between embryos resulting in positive and negative live birth. Embryo score was related to the odds of implantation and live birth in the oocyte donation program (odds ratio (OR)=1.29; 95% CI [1.19-1.39]; P < 0.001 for implantation and OR = 1.26; 95% CI [1.16-1.36]; P < 0.001 for live birth) and in conventional treatments with autologous oocytes (OR = 1.38; 95% CI [1.24-1.54]; P < 0.001 for implantation and OR = 1.47; 95% CI [1.30-1.65]; P < 0.001 for live birth). There was no significant association of embryo score with implantation or live birth in treatments involving PGT-A. LIMITATIONS, REASONS FOR CAUTION: This study is limited by its retrospective nature. Further prospective randomized trials are required to confirm the clinical impact of these findings. The single-center design should be taken into account when considering the universal application of the model. WIDER IMPLICATIONS OF THE FINDINGS: Evidence of the clinical efficiency of automated embryo scoring for ranking embryos with different morphological grade and potential in order to achieve higher implantation and live birth rates may make it a decision support tool for embryologists when selecting blastocysts for embryo transfer. STUDY FUNDING/COMPETING INTEREST(S): This research has been funded by a grant from the Ministry of Science, Innovation and Universities FIS (PI21/00283) awarded to M.M. There are no competing interests to declare. TRIAL REGISTRATION NUMBER: N/A.

Will the introduction of automated ART laboratory systems render the majority of embryologists redundant?

IVF techniques have changed over time with the aim of improving clinical results. Today, embryology is facing a change common to most areas of medicine, the introduction of automation. The use of automated systems in the IVF laboratory is already happening, for example, with electronic witnessing and the ranking of embryos according to their implantation potential. It is expected that in the near future, various systems in the IVF laboratory will be automated. In this way, gamete manipulation would cease to be manual and embryo culture and selection would be performed by means of microfluidics and artificial intelligence. Therefore, the tasks of the embryologist will inevitably be reduced. However, new functions related to data capture, management and analysis will emerge, along with other research skills and increased communication with other professionals and patients.

An artificial intelligence model based on the proteomic profile of euploid embryos and blastocyst morphology: a preliminary study.

RESEARCH QUESTION: The study aimed to develop an artificial intelligence model based on artificial neural networks (ANNs) to predict the likelihood of achieving a live birth using the proteomic profile of spent culture media and blastocyst morphology. DESIGN: This retrospective cohort study included 212 patients who underwent single blastocyst transfer at IVI Valencia. A single image of each of 186 embryos was studied, and the protein profile was analysed in 81 samples of spent embryo culture medium from patients included in the preimplantation genetic testing programme. The information extracted from the analyses was used as input data for the ANN. The multilayer perceptron and the back-propagation learning method were used to train the ANN. Finally, predictive power was measured using the area under the curve (AUC) of the receiver operating characteristic curve. RESULTS: Three ANN architectures classified most of the embryos correctly as leading (LB+) or not leading (LB-) to a live birth: 100.0% for ANN1 (morphological variables and two proteins), 85.7% for ANN2 (morphological variables and seven proteins), and 83.3% for ANN3 (morphological variables and 25 proteins). The artificial intelligence model using information extracted from blastocyst image analysis and concentrations of interleukin-6 and matrix metalloproteinase-1 was able to predict live birth with an AUC of 1.0. CONCLUSIONS: The model proposed in this preliminary report may provide a promising tool to select the embryo most likely to lead to a live birth in a euploid cohort. The accuracy of prediction demonstrated by this software may improve the efficacy of an assisted reproduction treatment by reducing the number of transfers per patient. Prospective studies are, however, needed.

Assessment of embryo implantation potential with a cloud-based automatic software.

RESEARCH QUESTION: Is embryo selection by Dana (automatic software for embryo evaluation) associated with a higher implantation rate in IVF treatments? DESIGN: A three-phase study for Dana system's validation: creation of a data-cloud of known implantation data (KID) embryos from 1676 transferred embryos; embryo evaluation by Dana considering manual annotations and embryo development videos (389 transferred embryos); and validation of Dana automatic selection, without embryologist's intervention (147 transferred embryos); RESULTS: The implantation rate of the 1021 KID embryos from phase 1 served to set four grades of embryos referring to implantation rate: A = 34%, B = 25%, C = 24%, and D = 19%. Phase 2: a classification ranking according to the unit average distance (UAD) and implantation potential was established: top (UAD ≤0.50), high (UAD = 0.51-0.66), medium (UAD = 0.67-1.03) and low (UAD >1.03). Pregnancy rates were 59%, 46%, 36% and 28%, respectively (P < 0.001). Phase 3: embryos were automatically categorized according to Dana's classification ranking. Most implanted embryos were found in groups top, high and medium (UAD ≤1.03), whereas the implantation rate in group low (UAD >1.03) was significantly lower: 46% versus 25%, respectively (P = 0.037). The twin gestation rate was higher when number of top embryos (UAD ≤0.5) transferred were two (52%) versus one (25%) (P < 0.001). CONCLUSIONS: Embryo selection based on Dana ranking increases the success of IVF treatments at least in oocyte donation programmes. The multicentre nature of the study supports its applicability at different clinics, standardizing the embryo development's interpretation. Dana's innovation is that the system increases its accuracy as the database grows.

Novel and conventional embryo parameters as input data for artificial neural networks: an artificial intelligence model applied for prediction of the implantation potential.

OBJECTIVE: To describe novel embryo features capable of predicting implantation potential as input data for an artificial neural network (ANN) model. DESIGN: Retrospective cohort study. SETTING: University-affiliated private IVF center. PATIENT(S): This study included 637 patients from the oocyte donation program who underwent single-blastocyst transfer during two consecutive years. INTERVENTION(S): None. MAIN OUTCOME MEASURE(S): The research was divided into two phases. Phase 1 consisted of the description and analysis of the following embryo features in implanted and nonimplanted embryos: distance and speed of pronuclear migration, blastocyst expanded diameter, inner cell mass area, and trophectoderm cell cycle length. Phase 2 consisted of the development of an ANN algorithm for implantation prediction. Results were obtained for four models fed with different input data. The predictive power was measured with the use of the area under the receiver operating characteristic curve (AUC). RESULT(S): Out of the five novel described parameters, blastocyst expanded diameter and trophectoderm cell cycle length had statistically different values in implanted and nonimplanted embryos. After the ANN models were trained and validated using fivefold cross-validation, they were capable of predicting implantation on testing data with AUCs of 0.64 for ANN1 (conventional morphokinetics), 0.73 for ANN2 (novel morphodynamics), 0.77 for ANN3 (conventional morphokinetics + novel morphodynamics), and 0.68 for ANN4 (discriminatory variables from statistical test). CONCLUSION(S): The novel proposed embryo features affect the implantation potential, and their combination with conventional morphokinetic parameters is effective as input data for a predictive model based on artificial intelligence.

Sperm chromosomal abnormalities and their contribution to human embryo aneuploidy.

In this work we reviewed 18 years of experience using fluorescence in situ hybridization (FISH) for sperm aneuploidy testing. We evaluated parameters associated with increased numerical sperm chromosome abnormalities and determined the male contribution to embryo aneploidies in terms of reproductive outcome by increased sperm aneuploidy. This retrospective study analyzed data from 2008 sperm samples of infertile males undergoing FISH analysis because of clinical history of repetitive implantation failure, recurrent miscarriage, impaired sperm parameters, or mixed causes. Sperm concentration was the only sperm parameter associated with FISH results-we observed a gradual increase of abnormal sperm FISH results in males with decreasing sperm concentration. However, a great proportion of normozoospermic males also showed increased sperm aneuploidies, suggesting that sperm parameters alone do not enable identification of a substantial proportion of infertile males at risk of sperm aneuploidies. Regarding reproductive outcomes, couples with normal sperm FISH results for the male had similar outcomes regardless of conventional in vitro fertilization (IVF), intracytoplasmic sperm injection (ICSI), or preimplantation genetic testing for aneuploidies (PGT-A). However, couples with abnormal sperm FISH results for the male showed better clinical outcomes after PGT-A, suggesting a potential contribution of sperm to embryo aneuploidy. Moreover, PGT-A cycles showed better clinical outcomes when 24 chromosomes were analyzed by array comparative genome hybridization (aCGH) or next-generation sequencing (NGS) instead of only nine chromosomes analyzed by FISH. In conclusion, sperm FISH analysis offers clinical prognostic value to evaluate reproductive possibilities in infertile couples. Therefore, couples with abnormal sperm FISH results should be offered genetic counseling and presented with clinical options such as PGT-A.

Sperm kinematic, head morphometric and kinetic-morphometric subpopulations in the blue fox (Alopex lagopus).

This work provides information on the blue fox ejaculated sperm quality needed for seminal dose calculations. Twenty semen samples, obtained by masturbation, were analyzed for kinematic and morphometric parameters by using CASA-Mot and CASA-Morph system and principal component (PC) analysis. For motility, eight kinematic parameters were evaluated, which were reduced to PC1, related to linear variables, and PC2, related to oscillatory movement. The whole population was divided into three independent subpopulations: SP1, fast cells with linear movement; SP2, slow cells and nonoscillatory motility; and SP3, medium speed cells and oscillatory movement. In almost all cases, the subpopulation distribution by animal was significantly different. Head morphology analysis generated four size and four shape parameters, which were reduced to PC1, related to size, and PC2, related to shape of the cells. Three morphometric subpopulations existed: SP1: large oval cells; SP2: medium size elongated cells; and SP3: small and short cells. The subpopulation distribution differed between animals. Combining the kinematic and morphometric datasets produced PC1, related to morphometric parameters, and PC2, related to kinematics, which generated four sperm subpopulations - SP1: high oscillatory motility, large and short heads; SP2: medium velocity with small and short heads; SP3: slow motion small and elongated cells; and SP4: high linear speed and large elongated cells. Subpopulation distribution was different in all animals. The establishment of sperm subpopulations from kinematic, morphometric, and combined variables not only improves the well-defined fox semen characteristics and offers a good conceptual basis for fertility and sperm preservation techniques in this species, but also opens the door to use this approach in other species, included humans.