Examining the efficacy of six published time-lapse imaging embryo selection algorithms to predict implantation to demonstrate the need for the development of specific, in-house morphokinetic selection algorithms

The mammalian sperm factor phospholipase C zeta is critical for early embryo division and pregnancy in humans and mice

STUDY QUESTION

Are sperm phospholipase C zeta (PLCζ) profiles linked to the quality of embryogenesis and pregnancy?

SUMMARY ANSWER

Sperm PLCζ levels in both mouse and humans correlate with measures of ideal embryogenesis whereby minimal levels seem to be required to result in successful pregnancy.

WHAT IS KNOWN ALREADY

While causative factors underlying male infertility are multivariable, cases are increasingly associated with the efficacy of oocyte activation, which in mammals occurs in response to specific profiles of calcium (Ca2+) oscillations driven by sperm-specific PLCζ. Although sperm PLCζ abrogation is extensively linked with human male infertility where oocyte activation is deficient, less is clear as to whether sperm PLCζ levels or localization underlies cases of defective embryogenesis and failed pregnancy following fertility treatment.

STUDY DESIGN, SIZE, DURATION

A cohort of 54 couples undergoing fertility treatment were recruited at the assisted reproductive technology laboratory at the King Faisal Hospital and Research Centre, Riyadh, Kingdom of Saudi Arabia. The recruitment criteria for males was a minimum sperm concentration of 5×106 sperm/ml, while all female patients had to have at least five oocytes. Sperm PLCζ analysis was performed in research laboratories, while semen assessments were performed, and time-lapse morphokinetic data were obtained, in the fertility clinic as part of routine treatment. The CRISPR/Cas9 system was concurrently used to induce indels and single-nucleotide mutations within the Plcζ gene to generate strains of Plcζ mutant mice. Sperm PLCζ was evaluated using immunofluorescence and immunoblotting with an antibody of confirmed consistent specificity against PLCζ.

PARTICIPANTS/MATERIALS, SETTING, METHODS

We evaluated PLCζ profiles in sperm samples from 54 human couples undergoing fertility treatment in the context of time-lapse morphokinetic analysis of resultant embryos, correlating such profiles to pregnancy status. Concurrently, we generated two strains of mutant Plcζ mice using CRISPR/Cas9, and performed IVF with wild type (WT) oocytes and using WT or mutant Plcζ sperm to generate embryos. We also assessed PLCζ status in WT and mutant mice sperm in the context of time-lapse morphokinetic analysis and breeding outcomes.

MAIN RESULTS AND THE ROLE OF CHANCE

A significant (P ≤ 0.05) positive relationship was observed between both PLCζ relative fluorescence and relative density with the times taken for both the second cell division (CC2) (r = 0.26 and r = 0.43, respectively) and the third cell division (S2) (r = 0.26). Examination of localization patterns also indicated significant correlations between the presence or absence of sperm PLCζ and CC2 (r = 0.27 and r = -0.27, respectively; P ≤ 0.025). Human sperm PLCζ levels were at their highest in the ideal times of CC2 (8-12 h) compared to time ranges outside the ideal timeframe (<8 and >12 h) where levels of human sperm PLCζ were lower. Following assignment of PLCζ level thresholds, quantification revealed a significantly higher (P ≤ 0.05) rate of successful pregnancy in values larger than the assigned cut-off for both relative fluorescence (19% vs 40%, respectively) and relative density (8% vs 54%, respectively). Immunoblotting indicated a single band for PLCζ at 74 kDa in sperm from WT mice, while a single band was also observed in sperm from heterozygous of Plcζ mutant mouse sperm, but at a diminished intensity. Immunofluorescent analysis indicated the previously reported (Kashir et al., 2021) fluorescence patterns in WT sperm, while sperm from Plcζ mutant mice exhibited a significantly diminished and dispersed pattern at the acrosomal region of the sperm head. Breeding experiments indicated a significantly reduced litter size of mutant Plcζ male mice compared to WT mice, while IVF-generated embryos using sperm from mutant Plcζ mice exhibited high rates of polyspermy, and resulted in significantly reduced numbers of these embryos reaching developmental milestones.

LIMITATIONS, REASONS FOR CAUTION

The human population examined was relatively small, and should be expanded to examine a larger multi-centre cohort. Infertility conditions are often multivariable, and it was not possible to evaluate all these in human patients. However, our mutant Plcζ mouse experiments do suggest that PLCζ plays a significant role in early embryo development.

WIDER IMPLICATIONS OF THE FINDINGS

We found that minimal levels of PLCζ within a specific range were required for optimal early embryogenesis, correlating with increased pregnancy. Levels of sperm PLCζ below specific thresholds were associated with ineffective embryogenesis and lower pregnancy rates, despite eliciting successful fertilization in both mice and humans. To our knowledge, this represents the first time that PLCζ levels in sperm have been correlated to prognostic measures of embryogenic efficacy and pregnancy rates in humans. Our data suggest for the first time that the clinical utilization of PLCζ may stand to benefit not just a specific population of male infertility where oocyte activation is completely deficient (wherein PLCζ is completely defective/abrogated), but also perhaps the larger population of couples seeking fertility treatment.

STUDY FUNDING/COMPETING INTEREST(S)

J.K. is supported by a faculty start up grant awarded by Khalifa University (FSU-2023-015). This study was also supported by a Healthcare Research Fellowship Award (HF-14-16) from Health and Care Research Wales (HCRW) to J.K., alongside a National Science, Technology, and Innovation plan (NSTIP) project grant (15-MED4186-20) awarded by the King Abdulaziz City for Science and Technology (KACST) for J.K. and A.M.A. The authors declare no conflicts of interest.

TRIAL REGISTRATION NUMBER

N/A.

A novel machine-learning framework based on early embryo morphokinetics identifies a feature signature associated with blastocyst development

BACKGROUND

Artificial Intelligence entails the application of computer algorithms to the huge and heterogeneous amount of morphodynamic data produced by Time-Lapse Technology. In this context, Machine Learning (ML) methods were developed in order to assist embryologists with automatized and objective predictive models able to standardize human embryo assessment. In this study, we aimed at developing a novel ML-based strategy to identify relevant patterns associated with the prediction of blastocyst development stage on day 5.

METHODS

We retrospectively analysed the morphokinetics of 575 embryos obtained from 80 women who underwent IVF at our Unit. Embryo morphokinetics was registered using the Geri plus® time-lapse system. Overall, 30 clinical, morphological and morphokinetic variables related to women and embryos were recorded and combined. Some embryos reached the expanded blastocyst stage on day 5 (BL Group, n = 210), some others did not (nBL Group, n = 365).

RESULTS

The novel EmbryoMLSelection framework was developed following four-steps: Feature Selection, Rules Extraction, Rules Selection and Rules Evaluation. Six rules composed by a combination of 8 variables were finally selected, and provided a predictive power described by an AUC of 0.84 and an accuracy of 81%.

CONCLUSIONS

We provided herein a new feature-signature able to identify with an high performance embryos with the best developmental competence to reach the expanded blastocyst stage on day 5. Clear and clinically relevant cut-offs were identified for each considered variable, providing an objective tool for early embryo developmental assessment.

Assessment of artificial intelligence model and manual morphokinetic annotation system as embryo grading methods for successful live birth prediction: a retrospective monocentric study

PURPOSE

The introduction of the time-lapse monitoring system (TMS) and the development of predictive algorithms could contribute to the optimal embryos selection for transfer. Therefore, the present study aims at investigating the efficiency of KIDScore and iDAScore systems for blastocyst stage embryos in predicting live birth events.

METHODS

The present retrospective study was conducted in a private IVF Unit setting throughout a 10-month period from October 2021 to July 2022, and included the analysis of 429 embryos deriving from 91 IVF/ICSI cycles conducted due to infertility of various etiologies. Embryos incubated at the Embryoscope+ timelapse incubator were analyzed through the established scoring systems: KIDScore and iDAScore®. The main outcome measure was the comparison of the two scoring systems in terms of live birth prediction. Embryos with the higher scores at day 5 (KID5 score/iDA5 score) were transferred or cryopreserved for later use.

RESULTS

Embryos with high KID5 and iDA5 scores positively correlated with the probability of successful live birth, with KID5 score yielding a higher efficiency in predicting a successful reproductive outcome compared to a proportionally high iDA5 score. KID5 demonstrated conservative performance in successfully predicting live birth compared to iDA5 score, indicating that an efficient prediction can be either provided by a relatively lower KID5 score or a relatively higher iDA5 score.

CONCLUSION

The developed artificial intelligence tools should be implemented in clinical practice in conjunction with the conventional morphological assessment for the conduction of optimized embryo transfer in terms of a successful live birth.

Morphokinetic Profiling Suggests That Rapid First Cleavage Division Accurately Predicts the Chances of Blastulation in Pig In Vitro Produced Embryos

The study of pig preimplantation embryo development has several potential uses: from agriculture to the production of medically relevant genetically modified organisms and from rare breed conservation to acting as a physiologically relevant model for progressing human and other (e.g., endangered) species' in vitro fertilisation technology. Despite this, barriers to the widespread adoption of pig embryo in vitro production include lipid-laden cells that are hard to visualise, slow adoption of contemporary technologies such as the use of time-lapse incubators or artificial intelligence, poor blastulation and high polyspermy rates. Here, we employ a commercially available time-lapse incubator to provide a comprehensive overview of the morphokinetics of pig preimplantation development for the first time. We tested the hypotheses that (a) there are differences in developmental timings between blastulating and non-blastulating embryos and (b) embryo developmental morphokinetic features can be used to predict the likelihood of blastulation. The abattoir-derived oocytes fertilised by commercial extended semen produced presumptive zygotes were split into two groups: cavitating/blastulating 144 h post gamete co-incubation and those that were not. The blastulating group reached the 2-cell and morula stages significantly earlier, and the time taken to reach the 2-cell stage was identified to be a predictive marker for blastocyst formation. Reverse cleavage was also associated with poor blastulation. These data demonstrate the potential of morphokinetic analysis in automating and upscaling pig in vitro production through effective embryo selection.

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.

A comparison of morphokinetic models and morphological selection for prioritizing euploid embryos: a multicentre cohort study

STUDY QUESTION

Are morphokinetic models better at prioritizing a euploid embryo for transfer over morphological selection by an embryologist?

SUMMARY ANSWER

Morphokinetic algorithms lead to an improved prioritization of euploid embryos when compared to embryologist selection.

WHAT IS KNOWN ALREADY

PREFER (predicting euploidy for embryos in reproductive medicine) is a previously published morphokinetic model associated with live birth and miscarriage. The second model uses live birth as the target outcome (LB model).

STUDY DESIGN, SIZE, DURATION

Data for this cohort study were obtained from 1958 biopsied blastocysts at nine IVF clinics across the UK from January 2021 to December 2022.

PARTICIPANTS/MATERIALS, SETTING, METHODS

The ability of the PREFER and LB models to prioritize a euploid embryo was compared against arbitrary selection and the prediction of four embryologists using the timelapse video, blinded to the morphokinetic time stamp. The comparisons were made using calculated percentages and normalized discounted cumulative gain (NDCG), whereby an NDCG score of 1 would equate to all euploid embryos being ranked first. In arbitrary selection, the ploidy status was randomly assigned within each cycle and the NDGC calculated, and this was then repeated 100 times and the mean obtained.

MAIN RESULTS AND THE ROLE OF CHANCE

Arbitrary embryo selection would rank a euploid embryo first 37% of the time, embryologist selection 39%, and the LB and PREFER ploidy morphokinetic models 46% and 47% of the time, respectively. The AUC for LB and PREFER model was 0.62 and 0.63, respectively. Morphological selection did not significantly improve the performance of both morphokinetic models when used in combination. There was a significant difference between the NDGC metric of the PREFER model versus embryologist selection at 0.96 and 0.87, respectively (t = 14.1, P < 0.001). Similarly, there was a significant difference between the LB model and embryologist selection with an NDGC metric of 0.95 and 0.87, respectively (t = 12.0, P < 0.001). All four embryologists ranked embryos similarly, with an intraclass coefficient of 0.91 (95% CI 0.82-0.95, P < 0.001).

LIMITATIONS, REASONS FOR CAUTION

Aside from the retrospective study design, limitations include allowing the embryologist to watch the time lapse video, potentially providing more information than a truly static morphological assessment. Furthermore, the embryologists at the participating centres were familiar with the significant variables in time lapse, which could bias the results.

WIDER IMPLICATIONS OF THE FINDINGS

The present study shows that the use of morphokinetic models, namely PREFER and LB, translates into improved euploid embryo selection.

STUDY FUNDING/COMPETING INTEREST(S)

This study received no specific grant funding from any funding agency in the public, commercial or not-for-profit sectors. Dr Alison Campbell is minor share holder of Care Fertility. All other authors have no conflicts of interest to declare. Time lapse is a technology for which patients are charged extra at participating centres.

TRIAL REGISTRATION NUMBER

N/A.

The Role of Artificial Intelligence and Machine Learning in Assisted Reproductive Technologies

Artificial intelligence (AI) and machine learning, the form most commonly used in medicine, offer powerful tools utilizing the strengths of large data sets and intelligent algorithms. These systems can help to revolutionize delivery of treatments, access to medical care, and improvement of outcomes, particularly in the realm of reproductive medicine. Whether that is more robust oocyte and embryo grading or more accurate follicular measurement, AI will be able to aid clinicians, and most importantly patients, in providing the best possible and individualized care. However, despite all of the potential strengths of AI, algorithms are not immune to bias and are vulnerable to the many socioeconomic and demographic biases that current healthcare systems suffer from. Wrong diagnoses as well is furthering of healthcare discrimination are real possibilities if both the capabilities and limitations of AI are not well understood. Armed with appropriate knowledge of how AI can most appropriately operate within medicine, and specifically reproductive medicine, will enable clinicians to both create and utilize machine learning-based innovations for the furthering of reproductive medicine and ultimately achieving the goal of building of healthy families.

An artificial intelligence algorithm for automated blastocyst morphometric parameters demonstrates a positive association with implantation potential

Blastocyst selection is primarily based on morphological scoring systems and morphokinetic data. These methods involve subjective grading and time-consuming techniques. Artificial intelligence allows for objective and quick blastocyst selection. In this study, 608 blastocysts were selected for transfer using morphokinetics and Gardner criteria. Retrospectively, morphometric parameters of blastocyst size, inner cell mass (ICM) size, ICM-to-blastocyst size ratio, and ICM shape were automatically measured by a semantic segmentation neural network model. The model was trained on 1506 videos with 102 videos for validation with no overlap between the ICM and trophectoderm models. Univariable logistic analysis found blastocyst size and ICM-to-blastocyst size ratio to be significantly associated with implantation potential. Multivariable regression analysis, adjusted for woman age, found blastocyst size to be significantly associated with implantation potential. The odds of implantation increased by 1.74 for embryos with a blastocyst size greater than the mean (147 ± 19.1 μm). The performance of the algorithm was represented by an area under the curve of 0.70 (p < 0.01). In conclusion, this study supports the association of a large blastocyst size with higher implantation potential and suggests that automatically measured blastocyst morphometrics can be used as a precise, consistent, and time-saving tool for improving blastocyst selection.

Using the embryo-uterus statistical model to predict pregnancy chances by using cleavage stage morphokinetics and female age: two centre-specific prediction models and mutual validation

BACKGROUND

The predictive capability of time-lapse monitoring (TLM) selection algorithms is influenced by patient characteristics, type and quality of data included in the analysis and the used statistical methods. Previous studies excluded DET cycles of which only one embryo implanted, introducing bias into the data. Therefore, we wanted to develop a TLM prediction model that is able to predict pregnancy chances after both single- and double embryo transfer (SET and DET).

METHODS

This is a retrospective study of couples (n = 1770) undergoing an in vitro fertilization cycle at the Erasmus MC, University Medical Centre Rotterdam (clinic A) or the Reinier de Graaf Hospital (clinic B). This resulted in 2058 transferred embryos with time-lapse and pregnancy outcome information. For each dataset a prediction model was established by using the Embryo-Uterus statistical model with the number of gestational sacs as the outcome variable. This process was followed by cross-validation.

RESULTS

Prediction model A (based on data of clinic A) included female age, t3-t2 and t5-t4, and model B (clinic B) included female age, t2, t3-t2 and t5-t4. Internal validation showed overfitting of model A (calibration slope 0.765 and area under the curve (AUC) 0.60), and minor overfitting of model B (slope 0.915 and AUC 0.65). External validation showed that model A was capable of predicting pregnancy in the dataset of clinic B with an AUC of 0.65 (95% CI: 0.61-0.69; slope 1.223, 95% CI: 0.903-1.561). Model B was less accurate in predicting pregnancy in the dataset of clinic A (AUC 0.60, 95% CI: 0.56-0.65; slope 0.671, 95% CI: 0.422-0.939).

CONCLUSION

Our study demonstrates a novel approach to the development of a TLM prediction model by applying the EU statistical model. With further development and validation in clinical practice, our prediction model approach can aid in embryo selection and decision making for SET or DET.

Sperm Chromatin Condensation Defect Accelerates the Kinetics of Early Embryonic Development but Does Not Modify ICSI Outcome

The origin and quality of gametes are likely to influence the kinetics of embryonic development. The purpose of the study was to assess the impact of sperm nuclear quality, and in particular sperm chromatin condensation, on the kinetics of early embryo development after intracytoplasmic sperm injection (ICSI). Our study included 157 couples who benefitted from ICSI for male factor infertility. Chromatin condensation and DNA fragmentation were assessed in spermatozoa prior to ICSI. Above the 20% threshold of sperm condensation defect, patients were included in the abnormal sperm chromatin condensation (ASCC) group; below the 20% threshold, patients were included in the normal sperm chromatin condensation (NSCC) group. After ICSI, the oocytes were placed in the time-lapse incubator. The kinetics of the cohort's embryonic development have been modeled. The fading times of pronuclei and the time to two blastomeres (t2, first cleavage) and four blastomeres (t4, third cleavage) differed significantly between the NSCC and ASCC groups, with earlier events occurring in the ASCC group. On the other hand, the state of sperm chromatin condensation did not seem to have an impact on live birth rates or the occurrence of miscarriages. The kinetics of early embryonic development was accelerated in males with a sperm chromatin condensation defect without compromising the chances of pregnancy or promoting miscarriage. However, our study highlights the paternal contribution to early embryonic events and potentially to the future health of the conceptus.

Zygote Diameter and Total Cytoplasmic Volume as Useful Predictive Tools of Blastocyst Quality

Introduction According to the Embryo Protection Act, the selection of embryos with the greatest potential for successful implantation in Germany must be performed in the pronucleus stage. The main aim of this study was to identify morphokinetic parameters that could serve as noninvasive biomarkers of blastocyst quality in countries with restrictive reproductive medicine laws. Materials and Methods The sample comprised 191 embryos from 40 patients undergoing antagonist cycles for intracytoplasmic sperm injection. Blastocysts were cultured in an EmbryoScope chamber and video records were validated to determine the post-injection timing of various developmental stages, cleavage stages, and blastocyst formation. The Gardner and Schoolcraft scoring system was used to characterize blastocyst quality. Results Morphokinetic data showed that the zygote diameter and total cytoplasmic volume were significantly different between good and poor blastocysts quality groups, where zygotes, which formed better blastocyst quality, had smaller diameter and smaller total cytoplasmic volume. Zygotes with more rapid pronuclear disappearance developed in better-quality blastocysts. Differences between good- and poor-quality blastocysts were also observed for late-stage parameters and for the spatial arrangement of blastomere where tetrahedral embryos more frequently forming good-quality blastocyst compare to the non-tetrahedral. Conclusions The study findings could be used to enhance embryo selection, especially in countries with strict Embryo Law Regulations. Further studies, including those in which the implantation potential and pregnancy rate are considered, are warranted to confirm these preliminary results.

Development of a dynamic machine learning algorithm to predict clinical pregnancy and live birth rate with embryo morphokinetics

Objective

Design

Setting

Patient(s)

Intervention(s)

Main Outcome Measure(s)

Result(s)

Conclusion(s)

Marginal differences in preimplantation morphokinetics between conventional IVF and ICSI in patients with preimplantation genetic testing for aneuploidy (PGT-A): A sibling oocyte study

Objective

This study aimed to analyze the morphokinetic behaviour between conventional IVF and ICSI, in cycles with preimplantation genetic testing for aneuploidies (PGT-A).

Materials

A randomized controlled trial (NCT03708991) was conducted in a private fertility center. Thirty couples with non-male factor infertility were recruited between November 2018 and April 2019. A total of 568 sibling cumulus oocyte complexes were randomly inseminated with conventional IVF and ICSI and cultured in an Embryoscope time-lapse system. The morphokinetic behaviour of IVF/ICSI sibling oocytes was analysed as primary endpoint. As secondary endpoints, morphokinetic parameters that predict blastocysts that will be biopsied, the day of biopsy, gender and euploid outcome was assessed.

Results

When comparing IVF to ICSI, only the time to reach the 2-cell stage (t2) was significantly delayed for IVF embryos: OR: 1.282 [1.020–1.612], p = 0.033. After standardizing for tPNf (ct parameters), only Blast(tStartBlastulation-t2) remained significant: OR: 0.803 [0.648–0.994], p = 0.044. For the analysis of zygotes that will be biopsied on day 5/6 versus zygotes without biopsy, only early morphokinetic parameters were considered. All parameters were different in the multivariate model: ct2: OR: 0.840 [0.709–0.996], p = 0.045; ct6: OR: 0.943 [0.890–0.998], p = 0.043; cc2(t3-t2): OR: 1.148 [1.044–1.263], p = 0.004; cc3(t5-t3): OR: 1.177 [1.107–1.251], p<0.0001. When comparing the development between blastocysts biopsied on day 5 versus day 6, only three morphokinetic parameters were significant: cc2(t3-t2): OR: 1.394 [1.010–1.926], p = 0.044; ctBlastocyst: OR: 0.613 [0.489–0.768], p<0.0001 and ctExpandedBlastocyst: OR: 0.913 [0.868–0.960], p = 0.0004. Multivariate analysis of gender and ploidy did not reveal differences in morphokinetic behaviour.

Conclusion

Minor morphokinetic differences are observed between IVF and ICSI. Early in the development, distinct cleavage patterns are observed between embryos that will be biopsied or not.

The KIDScore™ D3 scoring system contributes to the prediction of embryonic development potential: A promising tool for screening high-quality embryos

Using the KIDScoreTM D3 (KID3) scoring system, day 3 embryos observed by time-lapse imaging (TLI) were scored to explore the predictive value of the KID scoring system on the developmental potential of embryos. The kinetic parameters of 477 normal fertilized embryos from 77 patients who underwent TLI in our hospital from January 2019 to June 2020 were evaluated by KID3, and the embryos were divided into five groups according to the scores for retrospective analysis of blastocyst formation. Additionally, the high-quality blastocyst formation rate, pregnancy rate and early abortion rate were analyzed via KID3 and traditional morphological assessments, and comparisons of differences among different ages were also performed. In the KID3 estimate, the blastocyst or high-quality blastocyst formation rate in the score 5 group was markedly higher than that in the score 1–4 groups. Blastocyst or high-quality blastocyst formation rates in the A group (the results of two evaluation tools indicated they were excellent embryos) and the B group (KID3: excellent embryos, traditional evaluation: not excellent embryos) were evidently increased in comparison with the C or D group (KID3: not excellent embryos, traditional evaluation: excellent embryo or not, respectively). Furthermore, the percentages of score 5 embryos, blastocyst and high-quality blastocyst formation rates for patients ≥ 35 years old were markedly decreased compared with those for patients < 34 years old, while the trends of nondiploid cleavage, multinucleation and asymmetric division were the opposite. Collectively, the KID3 scoring system may be a promising predictive tool for screening embryos with better developmental potential.

Robust and generalizable embryo selection based on artificial intelligence and time-lapse image sequences

Assessing and selecting the most viable embryos for transfer is an essential part of in vitro fertilization (IVF). In recent years, several approaches have been made to improve and automate the procedure using artificial intelligence (AI) and deep learning. Based on images of embryos with known implantation data (KID), AI models have been trained to automatically score embryos related to their chance of achieving a successful implantation. However, as of now, only limited research has been conducted to evaluate how embryo selection models generalize to new clinics and how they perform in subgroup analyses across various conditions. In this paper, we investigate how a deep learning-based embryo selection model using only time-lapse image sequences performs across different patient ages and clinical conditions, and how it correlates with traditional morphokinetic parameters. The model was trained and evaluated based on a large dataset from 18 IVF centers consisting of 115,832 embryos, of which 14,644 embryos were transferred KID embryos. In an independent test set, the AI model sorted KID embryos with an area under the curve (AUC) of a receiver operating characteristic curve of 0.67 and all embryos with an AUC of 0.95. A clinic hold-out test showed that the model generalized to new clinics with an AUC range of 0.60–0.75 for KID embryos. Across different subgroups of age, insemination method, incubation time, and transfer protocol, the AUC ranged between 0.63 and 0.69. Furthermore, model predictions correlated positively with blastocyst grading and negatively with direct cleavages. The fully automated iDAScore v1.0 model was shown to perform at least as good as a state-of-the-art manual embryo selection model. Moreover, full automatization of embryo scoring implies fewer manual evaluations and eliminates biases due to inter- and intraobserver variation.

Oocyte quality and embryo selection strategies: a review for the embryologists, by the embryologists

With the advance of assisted reproduction techniques, and the trend towards blastocyst culture and single embryo transfer, gamete and embryo assessment have gained greater importance in ART treatment. Embryo quality depends mainly on gamete quality and culture conditions. Oocyte maturity identification is necessary in order to plan fertilization timing. Mature oocytes at the metaphase II stage show a higher fertilization rate compared to immature oocytes. Morphology assessment is a critical yet challenging task that may serve as a good prognostic tool for future development and implantation potential if done effectively. Various grading systems have been suggested to assess embryos at pronuclear, cleavage, and blastocyst stages. By identifying the embryo with the highest implantation potential, it is possible to reduce the number of embryos transferred without compromising the chances of a successful pregnancy. Apart from the conventional morphology assessment, there are several invasive or non-invasive methods for embryo selection such as preimplantation genetic testing, morphokinetics, proteomics, metabolomics, oxygen consumption, and measurement of oxidative stress in culture medium. Morphokinetics is a method based on time-lapse technology and continuous monitoring of embryos. In this review, we aim to describe and compare the most effective and widely used methods for gamete and embryo assessment as well as embryo selection.

Time-Lapse Systems: A Comprehensive Analysis on Effectiveness

Time-lapse systems have quickly become a common feature of in vitro fertilization laboratories all over the world. Since being introduced over a decade ago, the alleged benefits of time-lapse technology have continued to grow, from undisturbed culture conditions and round the clock, noninvasive observations to more recent computer-assisted selection of embryos through the development of algorithms. Despite the global uptake of time-lapse technology, its real impact on clinical outcomes is still controversial. This review aims to explore the different features offered by time-lapse technology, discussing incubation, algorithms, artificial intelligence and the regulation of nonessential treatment interventions, while assessing evidence on whether any benefit is offered over conventional technology.

Shallow artificial networks with morphokinetic time‐lapse parameters coupled to ART data allow to predict live birth

Purpose

The purpose of this work was to construct shallow neural networks (SNN) using time‐lapse technology (TLT) from morphokinetic parameters coupled to assisted reproductive technology (ART) parameters in order to assist the choice of embryo(s) to be transferred with the highest probability of achieving a live birth (LB).

Methods

A retrospective observational single‐center study was performed, 654 cycles were included. Three SNN: multilayers perceptron (MLP), simple recurrent neuronal network (simple RNN) and long short term memory RNN (LSTM‐RNN) were trained with K‐fold cross‐validation to avoid sampling bias. The predictive power of SNNs was measured using performance scores as AUC (area under curve), accuracy, precision, Recall and F1 score.

Results

In the training data group, MLP and simple RNN provide the best performance scores; however, all AUCs were above 0.8. In the validating data group, all networks were equivalent with no performance scores difference and all AUC values were above 0.8.

Conclusion

Coupling morphokinetic parameters with ART parameters allows to SNNs to predict the probability of LB, and all SNNs seems to be efficient according to the performance scores. An automatic time recognition system coupled to one of these SNNs could allow a complete automation to choose the blastocyst(s) to be transferred.

Aneuploidy rates and morphokinetic parameters of embryos cultured in distinct culture media: a sibling oocyte study

STUDY QUESTION

Do embryos from sibling oocytes assigned to distinct single-step media culture systems demonstrate differences in early embryo development, morphokinectics or aneuploidy rates?

SUMMARY ANSWER

Embryo quality, morphokinetic parameters and aneuploidy rates from trophectoderm biopsy were similar between sibling embryos cultured in distinct media systems from the time of gamete isolation.

WHAT IS KNOWN ALREADY

Studies on the effect of commercially available embryo culture media systems have demonstrated inconsistent impact on human embryonic development, morphokinetics, aneuploidy rates and clinical outcomes. In addition, these studies have been primarily randomized at the level of the embryo or the patient to culture media.

STUDY DESIGN, SIZE, DURATION

Prospective sibling oocyte cohort derived from 200 subjects undergoing IVF at a tertiary academic medical center between February 2018 and November 2019.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Sibling oocytes were allocated to Global® or SAGE® media system based upon laterality of ovary from which they were retrieved. All embryos were cultured in a time-lapse incubator. Blastocysts underwent trophectoderm biopsy for preimplantation genetic testing for aneuploidy using next-generation sequencing.

MAIN RESULTS AND THE ROLE OF CHANCE

One hundred twenty-seven subjects (n = 127) had paired blastocysts for biopsy in each culture media system. There was no difference in top quality blastocyst formation (47.1 ± 31.0 vs 48.1 ± 27.2%; P = 0.87) nor aneuploidy rate (62.3 ± 34.0 vs 56.1 ± 34.4%; P = 0.07) for sibling embryos cultured in Global versus SAGE media system. Embryo morphokinetic parameters including time to each cell division from two cells (t2) to eight cells (t8), time to morula stage (tM), time to blastocele formation (tSB), time to fully formed blastocyst (tB) and time to expansion of the blastocyst (tEB) were similar between paired blastocysts from each culture media system.

LIMITATIONS, REASONS FOR CAUTION

Pregnancy outcomes and offspring health data were not available for analysis.

WIDER IMPLICATIONS OF THE FINDINGS

Commercially available culture media may not have a differential impact on embryo development and blastocyst aneuploidy rate when patient and stimulation-related factors are held constant.

STUDY FUNDING/COMPETING INTEREST(S)

There was no external funding for this study. C.H. is owner of a consultancy company, IVF Professionals, Chief Scientific Officer at Apricity, Executive Director at TMRW and co-owner and shareholder of Aria Fertility. She has received speaker fees, consulting fees and travel support from Cooper Surgical and Vitrolife. J.B. is an employee and shareholder of vitrolife.

TRIAL REGISTRATION NUMBER

N/A.

Blastocyst Prediction of Day-3 Cleavage-Stage Embryos Using Machine Learning

Background: Embryo selection plays an important role in the success of in vitro fertilization (IVF). However, morphological embryo assessment has a number of limitations, including the time required, lack of accuracy, and inconsistency. This study determined whether a machine learning-based model could predict blastocyst formation using day-3 embryo images.

Methods: Day-3 embryo images from IVF/intracytoplasmic sperm injection (ICSI) cycles performed at My Duc Phu Nhuan Hospital between August 2018 and June 2019 were retrospectively analyzed to inform model development. Day-3 embryo images derived from two-pronuclear (2PN) zygotes with known blastocyst formation data were extracted from the CCM-iBIS time-lapse incubator (Astec, Japan) at 67 hours post ICSI, and labeled as blastocyst/non-blastocyst based on results at 116 hours post ICSI. Images were used as the input dataset to train (85%) and validate (15%) the convolutional neural network (CNN) model, then model accuracy was determined using the training and validation dataset. The performance of 13 experienced embryologists for predicting blastocyst formation based on 100 day-3 embryo images was also evaluated.

Results: A total of 1,135 images were allocated into training ([Formula: see text] 967) and validation ([Formula: see text] 168) sets, with an even distribution for blastocyst formation outcome. The accuracy of the final model for blastocyst formation was 97.72% in the training dataset and 76.19% in the validation dataset. The final model predicted blastocyst formation from day-3 embryo images in the validation dataset with an area under the curve of 0.75 (95% confidence interval [CI] 0.69–0.81). Embryologists predicted blastocyst formation with the accuracy of 70.07% (95% CI 68.12%–72.03%), sensitivity of 87.04% (95% CI 82.56%–91.52%), and specificity of 30.93% (95% CI 29.35%–32.51%).

Conclusions: The CNN-based machine learning model using day-3 embryo images predicted blastocyst formation more accurately than experienced embryologists. The CNN-based model is a potential tool to predict additional IVF outcomes.

Focus on time-lapse analysis: blastocyst collapse and morphometric assessment as new features of embryo viability

The main goal of assisted reproductive technology (ART) is to achieve a healthy singleton live birth after the transfer of one embryo. A major objective of IVF scientists has always been to use adequate criteria for selecting the embryo for transfer according to its implantation potential. Indeed, embryo quality is usually assessed by evaluating visual morphology, which relies on the removal of the embryo from the incubator and might include inter- and intra-evaluator variation among embryologists. Recently, an advancement in embryo culture has taken place with the introduction of a new type of incubator with an integrated time-lapse monitoring system, which enables embryologists to analyse the dynamic events of embryo development from fertilization to blastocyst formation. This novel practice is rapidly growing and has been used in many IVF centres worldwide. Therefore, the main aim of this review is to present the benefits of time-lapse monitoring in a modern embryology laboratory; in particular, we discuss blastocyst collapse and morphometric blastocyst assessment, and analyse their association with embryo viability and implantation potential. In addition, we highlight preliminary studies involving artificial intelligence and machine learning models as non-invasive markers of clinical pregnancy.

Change in the Strategy of Embryo Selection with Time-Lapse System Implementation-Impact on Clinical Pregnancy Rates

Time-lapse systems (TLS) and associated algorithms are interesting tools to improve embryo selection. This study aimed to evaluate how TLS and KIDScore™ algorithm changed our practices of embryo selection, as compared to a conventional morphological evaluation, and improved clinical pregnancy rates (CPR). In the study group (year 2020, n = 303 transfers), embryos were cultured in an EmbryoScope+ time-lapse incubator. A first team observed embryos conventionally once a day, while a second team selected the embryos for transfer based on time-lapse recordings. In the control group (year 2019, n = 279 transfers), embryos were selected using the conventional method, and CPR were recorded. In 2020, disagreement between TLS and the conventional method occurred in 32.1% of transfers, more often for early embryos (34.7%) than for blastocysts (20.5%). Irregular morphokinetic events (direct or reverse cleavage, multinucleation, abnormal pronuclei) were detected in 54.9% of the discordant embryos. When it was available, KIDScore™ was decreased for 73.2% of the deselected embryos. Discordant blastocysts mainly corresponded with a decrease in KIDScore™ (90.9%), whereas discordant Day 3 embryos resulted from a decreased KIDScore™ and/or an irregular morphokinetic event. CPR was significantly improved in the TLS group (2020), as compared to the conventional group (2019) (32.3% vs. 21.9%, p = 0.005), even after multivariate analysis. In conclusion, TLS is useful to highlight some embryo development abnormalities and identify embryos with the highest potential for pregnancy.

Comparing prediction of ongoing pregnancy and live birth outcomes in patients with advanced and younger maternal age patients using KIDScore™ day 5: a large-cohort retrospective study with single vitrified-warmed blastocyst transfer

BACKGROUND

The KIDScore™ Day 5 (KS-D5) model, version 3, is a general morphokinetic prediction model (Vitrolife, Sweden) for fetal heartbeat prediction after embryo transfer that was developed based on a large data set that included implantation results from a range of clinics with different patient populations, culture conditions and clinical practices. However, there was no study to comparing their pregnancy and live birth prediction ability among different maternal age. The aim of this study is to analyze the performance of KS-D5 in predicting pregnancy and live birth in various maternal age groups after single vitrified-warmed blastocyst transfer (SVBT).

METHODS

A total of 2486 single vitrified-warmed blastocyst transfer (SVBT) cycles were analyzed retrospectively. Confirmed fetal heartbeat positive (FHB+) and live birth (LB+) rates were stratified by Society for Assisted Reproductive Technology (SART) maternal age criteria (< 35, 35-37, 38-40, 41-42 and ≥ 43 years of age). Within each age group, the performance of the prediction model was calculated using the AUC, and the results were compared across the age groups.

RESULTS

In all age groups, the FHB+ rates decreased as the KIDScore decreased (P <  0.05). Conversely, the AUCs increased as the maternal age increased. The AUC of the < 35 age group (0.589) was significantly lower than the AUCs of the 41-42 age group (0.673) and the ≥43 age group (0.737), respectively (P <  0.05). In all age groups, the LB+ rates decreased as the KIDScore decreased (P <  0.05). Conversely, the AUCs increased as the maternal age increased. The AUC of the ≥43 age group (0.768) was significantly higher than the AUCs of other age groups (P <  0.05; < 35 age group = 0.596, 35-37 age group = 0.640, 38-40 age group = 0.646, 41-42 age group = 0.679).

CONCLUSIONS

In the present study, we determined that the KIDScore model worked well for prediction of pregnancy and live birth outcomes in advanced age patients.

A novel embryo quality scoring system to compare groups of embryos at different developmental stages

PURPOSE

To construct a new embryonic quality scoring system to compare groups of embryos at different developmental stages.

METHODS

Based on a hypothesis that the implantation potential of any embryo in an ovum pickup (OPU) cycle remains the same at any stage of development, be it day 2, 3, or 5, a new embryo quality scoring (EQS) system was designed. It was based on the analysis of the clinical results of 1610 single embryo transfers. We validated this scoring system in the comparison of embryonic quality between groups by evaluating the mean scores calculated at day 2, day 3, and day 5 for 957 embryos (150 cycles) from 3 different groups. We then compared EQSs of patients with pregnancy favorable factors (group A) such as young age and high AMH levels, with the patients with contra features (group B).

RESULTS

We confirmed that each mean EQS assessed at different stages of embryonic development within the same group was similar. The mean EQSs on day 3 and day 5 in group A were significantly higher than the mean EQSs on days 2, 3, and 5 in group B.

CONCLUSION

The novel EQS system proposed by us enables embryonic quality comparison between groups of embryos at different developmental stages.

Improving embryo selection by the development of a laboratory-adapted time-lapse model

OBJECTIVE

To study whether a powerful, in-house, embryo-selection model can be developed for a specific in vitro fertilization (IVF) laboratory where embryos were already selected for transfer using general models.

DESIGN

In total, 12,944 fertilized oocytes were incubated in an EmbryoScope (Vitrolife, Göteborg, Sweden) at our laboratory. Embryos were selected for transfer or freezing using general models. There were 1,879 embryos with known implantation data (KID), of which 425 had positive KIDs. For the outcome, we set 3 endpoints for KID's definition: gestational sac, clinical pregnancy, and live birth. Results of a comparison between KID-positive and -negative embryos for cell division timings were analyzed separately for intracytoplasmic sperm injection (ICSI) and IVF embryos in patients aged 18-41 years.

SETTING

IVF center.

PATIENTS

The study included 1,075 women undergoing IVF or ICSI treatment between June 2013 and February 2019.

INTERVENTION(S)

None.

MAIN OUTCOME MEASURE(S)

The KID-positive and -negative embryos were analyzed for statistical differences in cell division timing and cell cycle intervals. We used the EmbryoScope Stats software (Unisense FertiliTech, Aarhus, Denmark) for model development. The statistically different timing parameters were tested for their contribution to scoring in the model. The algorithms were tested for area under the receiver operating characteristic curve (AUC) in the KID embryos for developing day-2, -3, and -5 embryo-selection models. The validation of these algorithms was performed using calibration/validation procedures.

RESULTS

Because significant differences in morphokinetics were found between the KID-positive and KID-negative embryos in our laboratory, it was possible to use our specific KID data to develop an in-house model. The algorithms were developed for embryo selection on days 2, 3, and 5 in the ICSI embryos. In most cases, AUC was >0.65, which indicated that these models were valid in our laboratory. In addition, these AUC values were obtained from all gestational sac, clinical pregnancy, and live birth KID embryo databases tested. An increase in the predictability of the models was observed from days 2-3 to day 5 models. The AUC test results ranged between 0.657 and 0.673 for day 2 and day 3, respectively, and 0.803 for the day 5 model.

CONCLUSION

A model based on laboratory-specific morphokinetics was found to be complementary to general models and an important additive tool for improving single embryo selection. Developing an in-house laboratory-specific model requires many stages of sorting and characterization. Many insights were drawn about the model developing process. These may facilitate and improve the process in other laboratories.

Outcome of in vitro fertilization cycles with automatic time-lapse instrumentation combined with the early embryo viability assessment score

AIM

To evaluate whether embryo selection using the early embryo viability assessment (EEVA) score increases the ongoing pregnancy rate of in vitro fertilization and intracytoplasmic sperm injection patients.

METHODS

One hundred eighty-one patients whose serum anti mullerian hormone (AMH) level was greater than 0.5 ng/μL were enrolled in the study. All patients received oocyte retrieval repeatedly from June 2017 to January 2019. Transferred embryos were selected using the EEVA score and Veeck's criteria. We investigated the blastocyst rate according to the EEVA score and Veeck's criteria and also evaluated the clinical outcome following embryo transfer of the blastocysts.

RESULTS

Blastocyst development rate (48.7%) and high-quality blastocyst (42.4%) of Veeck 1 was statistically higher than others. The blastocyst rate (71.4%) and high-quality blastocyst rate (60.0%) for EEVA 1 was the highest, and a correlation between the EEVA score and the blastocyst rate was also identified in cases younger than 40 years. Blastocyst rate of EEVA 1 + 2 (69.8% 208/298) was statistically higher than that of Veeck 1 + 2 (40.1% 317/791) (p < 0.05) and high-quality blastocyst rate of EEVA 1 + 2 (50.0% 104/208) was also higher than that of Veeck 1 + 2 (36.6% 117/320) (p < 0.05). However, there was a significant correlation between EEVA and the pregnancy rate and pregnancy rate of EEVA 1 + 2 showed no statistical difference compared with Veeck 1 + 2.

CONCLUSIONS

Although it remains to be answered whether a computer can substitute Veeck's classification, the EEVA score could be a viable alternative to predict the blastocyst rate and to select those high-potential embryos that improve the pregnancy rate.

Development of deep learning algorithms for predicting blastocyst formation and quality by time-lapse monitoring

Approaches to reliably predict the developmental potential of embryos and select suitable embryos for blastocyst culture are needed. The development of time-lapse monitoring (TLM) and artificial intelligence (AI) may help solve this problem. Here, we report deep learning models that can accurately predict blastocyst formation and usable blastocysts using TLM videos of the embryo’s first three days. The DenseNet201 network, focal loss, long short-term memory (LSTM) network and gradient boosting classifier were mainly employed, and video preparation algorithms, spatial stream and temporal stream models were developed into ensemble prediction models called STEM and STEM⁺. STEM exhibited 78.2% accuracy and 0.82 AUC in predicting blastocyst formation, and STEM⁺ achieved 71.9% accuracy and 0.79 AUC in predicting usable blastocysts. We believe the models are beneficial for blastocyst formation prediction and embryo selection in clinical practice, and our modeling methods will provide valuable information for analyzing medical videos with continuous appearance variation.

Liao et al. propose a deep learning model to predict blastocyst formation using TLM videos following the first three days of embryogenesis. The authors develop an ensemble prediction model, STEM and STEM⁺, which were found to exhibit 78.2% and 71.9% accuracy at predicting blastocyst formation and useable blastocysts respectively.

A machine learning system with reinforcement capacity for predicting the fate of an ART embryo

The aim of this work was to construct a score issued from a machine learning system with self-improvement capacity able to predict the fate of an ART embryo incubated in a time lapse monitoring (TLM) system. A retrospective study was performed. For the training data group, 110 couples were included and, 891 embryos were cultured. For the global setting data group, 201 couples were included, and 1186 embryos were cultured. No image analysis was used; morphokinetic parameters from the first three days of embryo culture were used to perform a logistic regression between the cell number and time. A score named DynScore was constructed, the prediction power of the DynScore on blastocyst formation and the baby delivery were tested via the area under the curve (AUC) obtained from the receiver operating characteristic (ROC). In the training data group, the DynScore allowed the blastocyst formation prediction (AUC = 0.634, p < 0.001), this approach was the higher among the set of the tested scores. Similar results were found with the global setting data group (AUC = 0.638, p < 0.001) and it was possible to increase the AUC of the DynScore with a regular update of the prediction system by reinforcement, with an AUC able to reach a value above 0.9. As only the best blastocysts were transferred, none of the tested scores was able to predict delivery. In conclusion, the DynScore seems to be able to predict the fate of an embryo. The reinforcement of the prediction system allows maintaining the predictive capacity of DynScore irrespective of the various events that may occur during the ART process. The DynScore could be implemented in any TLM system and adapted by itself to the data of any ART center.Abbreviations: ART: assisted reproduction technology; TLM: time lapse monitoring system; AUC: area under the curve; ROC: receiver operating characteristic; eSET: elective single embryo transfer; AIS: artificial intelligence system; KID: known implantation data; AMH: anti-Müllerian hormone; BMI: body mass index; WHO: World Health Organization; c-IVF: conventional in-vitro fertilization; ICSI: intracytoplasmic sperm injection; PNf: pronuclear formation; D3: day 3; D5: day 5; D6: day 6; GnRH: gonadotrophin releasing hormone; FSH: follicle stimulating hormone; LH: luteinizing hormone; hCG: human chorionic gonadotropin; PVP: polyvinyl pyrrolidone; PNf: time of pronuclear fading; tx: time of cleavage to x blastomeres embryo; ICM: inner cell mass; TE: trophectoderm; NbCell_t: number of cells at t time; FIFO: first in first out; TD: training data group; SD: setting data group; R: real world.

The clinical use of time-lapse in human-assisted reproduction

A major challenge in the assisted reproduction laboratory is to set up reproducible and efficient criteria to identify the embryo with the highest developmental potential. Over the years, several methods have been used worldwide with this purpose. Initially, standard morphology assessment was the only available strategy. It is now universally recognized that besides being a very subjective embryo selection strategy, morphology evaluation alone has a very poor prognostic value. More recently, the availability of time-lapse incubators allowed a continuous monitoring of human embryo development. This technology has spread quickly and many fertility clinics over the world produced a remarkable amount of data. To date, however, a general consensus on which variables, or combination of variables, should play a central role in embryo selection is still lacking. Many confounding factors, concerning both patient features and clinical and biological procedures, have been observed to influence embryo development. In addition, several studies have reported unexpected positive outcomes, even in the presence of abnormal developmental criteria. While it does not seem that time-lapse technology is ready to entirely replace the more invasive preimplantation genetic testing in identifying the embryo with the highest implantation potential, it is certainly true that its application is rapidly growing, becoming progressively more accurate. Studies involving artificial intelligence and deep-learning models as well as combining morphokinetic with other non-invasive markers of embryo development, are currently ongoing, raising hopes for its successful applicability for clinical purpose in the near future. The present review mainly focuses on data published starting from the first decade of 2000, when time-lapse technology was introduced as a routine clinical practice in the infertility centers.

Use of time-lapse monitoring in medically assisted reproduction treatments: a mini-review

During human in vitro culture, a morphological microscope analysis is normally performed to select the best embryo to transfer, with the hope of obtaining a successful pregnancy. The morphological evaluation may combine number and size of blastomeres, fragmentation, multinucleation, blastocyst expansion, inner-cell mass and trophectoderm appearance. However, standard microscopy evaluation involves the removal of the embryos from the incubator, exposing them to changes in pH, temperature, and oxygen level. Additionally, morphological assessments might include high inter-observer variability. Recently, continuous embryo culture using time-lapse monitoring (TLM) has allowed embryologists to analyse the dynamic and morphokinetic events of embryo development and, based on that, the embryologist is able to scrutinize the complete sequence of embryonic evolution, from fertilization to the blastocyst formation. Therefore, TLM allows an uninterrupted culture condition, reducing the need to remove embryos from the incubator. The monitoring system is normally composed of a standard incubator with an integrated microscope coupled to a digital camera, which is able to collect images at regular times, and subsequently processed into video. These data can be annotated and analyzed using an integrated software, therefore this allows embryologists to facilitate the process of embryo selection for transfer. The main aim of this paper is to discuss the potential benefits and uses of the TLM in the embryology laboratory.

Can embryo morphokinetic parameters predict euploid pregnancy loss?

OBJECTIVE

To use time-lapse imaging to compare embryo morphokinetic parameters between embryos resulting in euploid pregnancy loss and euploid embryos resulting in live birth.

DESIGN

Retrospective cohort study.

SETTING

Single academic fertility center.

PATIENT(S)

All euploid single embryo transfers between October 2015 and January 2018.

INTERVENTION(S)

Collection and analysis of baseline characteristics, cycle parameters, and outcomes.

MAIN OUTCOME MEASURE(S)

Embryo morphokinetic measurements assessed with time-lapse imaging for time to syngamy (TPNf), time to two cells, time to three cells, time to four cells, time to eight cells, time to morula, and time to blastocyst.

RESULT(S)

The study included 192 euploid single-embryo transfers. Of these, the pregnancy rate was 78% (150 of 193) and the live-birth rate was 63% (121 of 193). There were 43 transfers that did not result in pregnancy, 15 biochemical pregnancy losses, 13 clinical losses, and 121 live births. There was no statistically significant difference in age, body mass index, or number of oocytes retrieved between the groups. Unadjusted and adjusted models revealed no differences in the morphokinetics of embryos resulting in euploid miscarriage compared with those resulting in live birth.

CONCLUSION(S)

Embryos that resulted in a euploid miscarriage did not display evidence of abnormal morphokinetics on time-lapse imaging. Euploid pregnancy loss is likely multifactorial, including both embryo and endometrial factors. Further research is needed to identify factors that can predict and prevent euploid loss.

Mining of variables from embryo morphokinetics, blastocyst's morphology and patient parameters: an approach to predict the live birth in the assisted reproduction service

Based on growing demand for assisted reproduction technology, improved predictive models are required to optimize in vitro fertilization/intracytoplasmatic sperm injection strategies, prioritizing single embryo transfer. There are still several obstacles to overcome for the purpose of improving assisted reproductive success, such as intra- and inter-observer subjectivity in embryonic selection, high occurrence of multiple pregnancies, maternal and neonatal complications. Here, we compare studies that used several variables that impact the success of assisted reproduction, such as blastocyst morphology and morphokinetic aspects of embryo development as well as characteristics of the patients submitted to assisted reproduction, in order to predict embryo quality, implantation or live birth. Thereby, we emphasize the proposal of an artificial intelligence-based platform for a more objective method to predict live birth.

Morphokinetics of early equine embryo development in vitro using time-lapse imaging, and use in selecting blastocysts for transfer

The use of time-lapse imaging (TLI) in the evaluation of morphokinetics associated with invitro developmental competence is well described for human, cattle and pig embryos. It is generally accepted that embryos that complete early cleavage sooner are more likely to form blastocysts and that timing of later events, such as blastocyst formation and expansion, are predictive of implantation potential and euploid status. In the horse, morphokinetics as a predictor of developmental competence has received little attention. In this study we evaluated the morphokinetics of early equine embryo development invitro for 144 oocytes after intracytoplasmic sperm injection and report the timings of blastocyst development associated with ongoing pregnancy for the first time. There was a tendency for time of cytoplasmic extrusion and first cleavage to occur earlier in the embryos that went on to form blastocysts (n=19) compared with those that arrested, and for first cleavage to occur earlier in blastocysts that established pregnancies that were ongoing (n=4) compared with pregnancies that were lost (n=2). TLI was clinically useful in identifying blastocysts when evaluation of morphology on static imaging was equivocal.

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.

Faster fertilization and cleavage kinetics reflect competence to achieve a live birth after intracytoplasmic sperm injection, but this association fades with maternal age

OBJECTIVE

To assess the relationship of early developmental kinetics with competence to provide a live birth and the impact of maternal age in this context.

DESIGN

Retrospective cohort study including 4,915 embryos, of which 1,390 were transferred and provided a clinical outcome paired with morphokinetic data; 168 of them resulted in a live birth (LB), and 1,222 did not (NLB). Early morphokinetic parameters were compared between LB and NLB embryos from patients stratified into two age groups (<37 and ≥37 years), and between embryos at the same competence group from patients aged <37 and ≥37 years. The association of morphokinetic parameters with live birth was tested by univariate and multivariate analyses.

SETTING

Fertility clinic.

PATIENT(S)

The study population included 1,066 patients undergoing autologous intracytoplasmic sperm injection cycles with fresh single (SET), double (DET) or triple (TET) embryo transfers on day 2 or 3. Of them, 669 patients produced NLB embryos and 134 produced LB embryos.

INTERVENTION(S)

None.

MAIN OUTCOME MEASURE(S)

Fertilization and cleavage morphokinetic parameters and live birth.

RESULT(S)

In the total patient population, all morphokinetic parameters were achieved earlier in LB compared with NLB embryos. The same was observed in patients aged <37 years (P<.015), but not ≥37 years. Except for the t8 (time at which an 8-blastomere embryo was identified), all morphokinetic parameters were reached earlier in LB embryos from patients aged <37 years compared with LB embryos from patients aged ≥37 years. Univariate analysis revealed that earlier occurrence of all morphokinetic parameters was associated with live birth, although only earlier t2 (time at which two separate and distinct cells were identified) was associated with live birth independently from maternal age in the multivariate analysis.

CONCLUSION(S)

Despite its retrospective nature and performance in a single IVF center, this study presents novel data indicating that embryos competent to provide a live birth display overall faster early developmental kinetics compared with embryos that do not achieve a live birth after transfer, a difference that, however, narrows as maternal age advances. The findings suggest that fertilization and cleavage morphokinetic parameters may constitute valuable references for embryo selection strategies aiming to improve live birth rates, specifically before advanced maternal age while holding limited usefulness in advanced maternal age.

Time-lapse imaging: the state of the art†

The introduction of time-lapse imaging to clinical in vitro fertilization practice enabled the undisturbed monitoring of embryos throughout the entire culture period. Initially, the main objective was to achieve a better embryo development. However, this technology also provided an insight into the novel concept of morphokinetics, parameters regarding embryo cell dynamics. The vast amount of data obtained defined the optimal ranges in the cell-cycle lengths at different stages of embryo development. This added valuable information to embryo assessment prior to transfer. Kinetic markers became part of embryo evaluation strategies with the potential to increase the chances of clinical success. However, none of them has been established as an international standard. The present work aims at describing new approaches into time-lapse: progress to date, challenges, and possible future directions.

Randomized study of G-TL and global media for blastocyst culture in the EmbryoScope: morphokinetics, pregnancy, and live births after single-embryo transfer

OBJECTIVE

To evaluate the efficacy of two different in vitro fertilization culture media for blastocyst development, pregnancy, and live birth rate. Global (GB) medium (used without refreshment) and G-TL medium (designed specifically for culture in time-lapse incubators) were compared.

DESIGN

Prospective randomized study of sibling embryo culture in two culture media.

SETTING

In vitro fertilization clinic.

PATIENT(S)

Women undergoing fresh or frozen cycles using autologous or donor oocytes.

INTERVENTION(S)

None.

MAIN OUTCOME MEASURE(S)

The primary endpoints were implantation, pregnancy, and live birth rate (LBR) after single blastocyst transfer. Secondary endpoints included embryo morphokinetics, development of good-quality blastocysts, and euploidy rate.

RESULT(S)

Kinetic data from 10,768 sibling pronucleate embryos cultured in the EmbryoScope were compared. GB embryos initiated compaction earlier and formed morula sooner than their G-TL counterparts. The mean timing for start of blastulation did not differ. The interval between start of blastulation and time of blastocyst formation was observed to be <12 hours for proportionately more GB compared with G-TL-cultured embryos. Despite a higher rate of observed dysmorphisms in GB embryos, the euploidy rate among biopsied blastocysts did not differ between media. A total of 820 single-embryo transfer cycles were performed. Implantation rates were similar between media, independent of whether the embryo transferred was fresh (GB 58.7% vs. G-TL 61.7%) or frozen (GB 64.1% vs. G-TL 60.5%). Live birth rates were also not different. With GB medium, the LBR for fresh and frozen transfers was 54.2% and 53.1%, respectively, as compared with 51.1% and 50%, respectively, with G-TL.

CONCLUSION(S)

Uninterrupted culture in a time-lapse incubator without medium refreshment was well supported by both media tested. Differences in morphokinetics did not necessarily dictate the superiority of one media over the other. Both pregnancy and LBR were not significantly influenced by choice of culture medium. The euploidy rate was also independent of culture medium.

Time of morulation and trophectoderm quality are predictors of a live birth after euploid blastocyst transfer: a multicenter study

OBJECTIVE

To investigate whether the morphodynamic characterization of a euploid blastocyst's development allows a higher prediction of a live birth after single-embryo-transfer (SET).

DESIGN

Observational cohort study conducted in two phases: training and validation.

SETTING

Private in vitro fertilization centers.

PATIENT(S)

Euploid blastocysts: 511 and 319 first vitrified-warmed SETs from 868 and 546 patients undergoing preimplantation genetic testing for aneuploidies (PGT-A) in the training and validation phase, respectively.

INTERVENTION(S)

Data collected from time of polar body extrusion to time of starting blastulation, and trophectoderm and inner-cell-mass static morphology in all embryos cultured in a specific time-lapse incubator with a continuous medium. Logistic regressions conducted to outline the variables showing a statistically significant association with live birth. In the validation phase, these variables were tested in an independent data set.

MAIN OUTCOME MEASURE(S)

Live births per SET.

RESULT(S)

The average live birth rate (LBR) in the training set was 40% (N = 207/511). Only time of morulation (tM) and trophectoderm quality were outlined as putative predictors of live birth at two IVF centers. In the validation set, the euploid blastocysts characterized by tM <80 hours and high-quality trophectoderm resulted in a LBR of 55.2% (n = 37/67), while those with tM ≥ 80 hours and a low-quality trophectoderm resulted in a LBR of 25.5% (N = 13/51).

CONCLUSION(S)

Time of morulation and trophectoderm quality are better predictors of a euploid blastocyst's reproductive competence. Our evidence was reproducible across different centers under specific culture conditions. These data support the crucial role of morulation for embryo development, a stage that involves massive morphologic, cellular, and molecular changes and deserves more investigation.

Time-lapse technology for embryo culture and selection

Culturing of human embryos in optimal conditions is crucial for a successful in vitro fertilisation (IVF) programme. In addition, the capacity to assess and rank embryos correctly for quality will allow for transfer of the potentially 'best' embryo first, thereby shortening the time to pregnancy, although not improving cumulative pregnancy and live birth rates. It will also encourage and facilitate the implementation of single embryo transfers, thereby increasing safety for mother and offspring. Time-lapse technology introduces the concept of stable culture conditions, in connection with the possibility of continuous viewing and documenting of the embryo throughout development. However, so far, even when embryo quality scoring is based on large datasets, or when using the time-lapse technology, the morphokinetic scores are still mainly based on subjective and intermittent annotations of morphology and timings. Also, the construction of powerful algorithms for widespread use is hampered by large variations in culture conditions between individual IVF laboratories. New methodology, involving machine learning, where every image from the time-lapse documentation is analysed by a computer programme, looking for patterns that link to outcome, may in the future provide a more accurate and non-biased embryo selection.

Impact of sperm characteristics on time-lapse embryo morphokinetic parameters and clinical outcome of conventional in vitro fertilization

BACKGROUND

Sperm abnormalities may negatively affect embryo development.

OBJECTIVES

To determine the influence of sperm abnormalities (morphology, motility, DNA fragmentation) on embryo morphokinetic variables and clinical outcome of conventional IVF.

MATERIALS AND METHODS

Participants were 86 couples undergoing in vitro fertilization (IVF). Sperm morphology was evaluated according to the strict criteria proposed by Kruger/Tygerberg. CASA system was applied for sperm motility assessment. Sperm DNA fragmentation was assessed by the chromatin structure assay (SCSA). Morphokinetic parameters were determined in 223 embryos obtained from conventional IVF only and cultured in a single-step medium using time-lapse imaging technology.

RESULTS

Time-lapse variables from the initial embryo development, such as time of pronuclei fading (tPNf) and time for two cells (t2), were those more strongly related with abnormalities of sperm motility, morphology, and DNA fragmentation. Sperm morphological abnormalities rather than sperm motility were more closely associated with embryo morphokinetics. Sperm head defects were mainly correlated with the last stages of embryonic development (t9 to tHB), sperm midpiece defects with intermediate cleaving embryos (t5-t9), and sperm tail defects with the initial stages of embryonic development (tPNa-t4). Excess residual cytoplasm was positively correlated with all embryo morphokinetic parameters except t2 and tM. Absence of acrosomes, pinheads, coiled tails, and multiple sperm morphological defects correlated negatively with time-lapse embryo morphokinetic variables.

DISCUSSION

A large number of sperm-related variables, including frequency of specific morphological defects, morphological indexes, DNA fragmentation and motility, and time-lapse embryo variables, such as time intervals based mainly of 15 time points were recorded.

CONCLUSION

There were strong associations between specific sperm defects of the head, midpiece, and tail with certain stages of embryonic development from observation of pronuclei to the hatched blastocyst. Coiled tail, cumulative head defects, and multiple abnormalities index (MAI) were associated both with embryo morphokinetics and the implantation success.

Development of automated annotation software for human embryo morphokinetics

STUDY QUESTION

Is it possible to develop an automated annotation tool for human embryo development in time-lapse devices based on image analysis?

SUMMARY ANSWER

We developed and validated an automated software for the annotation of human embryo morphokinetic parameters, having a good concordance with expert manual annotation on 701 time-lapse videos.

WHAT IS KNOWN ALREADY

Morphokinetic parameters obtained with time-lapse devices are increasingly used for the assessment of human embryo quality. However, their annotation is time-consuming and can be slightly operator-dependent, highlighting the need to develop fully automated approaches.

STUDY DESIGN, SIZE, DURATION

This monocentric study was conducted on 701 videos originating from 584 couples undergoing IVF with embryo culture in a time-lapse device. The only selection criterion was that the duration of the video must be over 60 h.

PARTICIPANTS/MATERIALS, SETTING, METHODS

An automated morphokinetic annotation tool was developed based on gray level coefficient of variation and detection of the thickness of the zona pellucida. The detection of cellular events obtained with the automated tool was compared with those obtained manually by trained experts in clinical settings.

MAIN RESULTS AND THE ROLE OF CHANCE

Although some differences were found when embryos were considered individually, we found an overall concordance between automated and manual annotation of human embryo morphokinetics from fertilization to expanded blastocyst stage (r2 = 0.92).

LIMITATIONS, REASONS FOR CAUTION

These results should undergo multicentric external evaluation in order to test the overall performance of the annotation tool. Getting access to the export of 3D videos would enhance the quality of the correlation with the same algorithm and its extension to the 3D regions of interest. A technical limitation of our work lies within the duration of the video. The more embryo stages the video contains, the more information the script has to identify them correctly.

WIDER IMPLICATIONS OF THE FINDINGS

Our system paves the way for high-throughput analysis of multicentric morphokinetic databases, providing new insights into the clinical value of morphokinetics as a predictor of embryo quality and implantation.

STUDY FUNDING/COMPETING INTEREST(S)

This study was partly funded by Finox-Gedeon Richter Forward Grant 2016 and NeXT (ANR-16-IDEX-0007). We have no conflict of interests to declare.

TRIAL REGISTRATION NUMBER

N/A

Intraindividual Embryo Morphokinetics Are Not Affected by a Switch of the Ovarian Stimulation Protocol Between GnRH Agonist vs. Antagonist Regimens in Consecutive Cycles

Background: The impact of controlled ovarian stimulation (COS) during medically assisted reproduction (MAR) on human embryogenesis is still unclear. Therefore, we investigated if early embryonic development is affected by the type of gonadotropin-releasing hormone (GnRH) analog used to prevent a premature LH surge. We compared embryo morphology and morphokinetics between GnRH agonist and antagonist cycles, both involving human chorionic gonadotropin (hCG)-trigger. To reduce possible confounding factors, we used intraindividual comparison of embryo morphokinetics in consecutive treatment cycles of the same patients that underwent a switch in the COS protocol. Methods: This retrospective cohort study analyzed morphokinetics of embryos from patients (n = 49) undergoing a switch in COS protocols between GnRH agonists followed by GnRH antagonists, or vice versa, after culture in a time-lapse incubator (EmbryoScope®, Vitrolife) in our clinic between 06/2011 and 11/2016 (n = 49 GnRH agonist cycles with n = 172 embryos; n = 49 GnRH antagonist cycles with n = 163 embryos). Among time-lapse cycles we included all embryos of the two consecutive cycles before and after a switch in the type of COS in the same patient. In-vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI) was performed and embryos were imaged up to day 5. Data were analyzed using Mann-Whitney U test or Fisher's exact test. The significance level was set to p = 0.05. Patients with preimplantation genetic screening cycles were excluded. Results: The mean age (years ± standard deviation) of patients at the time of treatment was 35.7 ± 4.3 (GnRH agonist) and 35.8 ± 4.0 (GnRH antagonist) (p = 0.94). There was no statistically significant difference in the number of oocytes collected or the fertilization rate. The numbers of top quality embryos (TQE), good-quality embryos (GQE), or poor-quality embryos (PQE) were also not different in GnRH agonist vs. antagonist cycles. We found no statistically significant difference between the analyzed morphokinetic parameters between the study groups. Conclusions: Our finding supports the flexible use of GnRH analogs to optimize patient treatment for COS without affecting embryo morphokinetics.

Performance of Day 5 KIDScore™ morphokinetic prediction models of implantation and live birth after single blastocyst transfer

PURPOSE

While several studies reported the association between morphokinetic parameters and implantation, few predictive models were developed to predict implantation after day 5 embryo transfer, generally without external validation. The objective of this study was to evaluate the respective performance of 2 commercially available morphokinetic-based models (KIDScore™ Day 5 versions 1 and 2) for the prediction of implantation and live birth after day 5 single blastocyst transfer.

METHODS

This monocentric retrospective study was conducted on 210 ICSI cycles with single day 5 embryo transfer performed with a time-lapse imaging (TLI) system between 2013 and 2016. The association between both KIDScore™ and the observed implantation and live birth rates was calculated, as well as the agreement between embryologist's choice for transfer and embryo ranking by the models.

RESULTS

Implantation and live birth rate were both 35.7%. A significant positive correlation was found between both models and implantation rate (r = 0.96 and r = 0.90, p = 0.01) respectively. Both models had statistically significant but limited predictive power for implantation (AUC 0.60). There was a fair agreement between the embryologists' choice and both models (78% and 61% respectively), with minor differences in case of discrepancies.

CONCLUSIONS

KIDScore™ Day 5 predictive models are significantly associated with implantation rates after day 5 single blastocyst transfer. However, their predictive performance remains perfectible. The use of these predictive models holds promises as decision-making tools to help the embryologist select the best embryo, ultimately facilitating the implementation of SET policy. However, embryologists' expertise remains absolutely necessary to make the final decision.