Correlations between a deep learning-based algorithm for embryo evaluation with cleavage-stage cell numbers and fragmentation.

RESEARCH QUESTION: Do cell numbers and degree of fragmentation in cleavage-stage embryos, assessed manually, correlate with evaluations made by deep learning algorithm model iDAScore v2.0? DESIGN: Retrospective observational study (n = 5040 embryos; 1786 treatments) conducted at two Swedish assisted reproductive technology centres between 2016 and 2021. Fresh single embryo transfer was carried out on days 2 or 3 after fertilization. Embryo evaluation using iDAScore v2.0 was compared with manual assessment of numbers of cells and grade of fragmentation, analysed by video sequences. RESULTS: Data from embryos transferred on days 2 and 3 showed that having three or fewer cells compared with four or fewer cells on day 2, and six or fewer cells versus seven to eight cells on day 3, correlated significantly with a difference in iDAScore (medians 2.4 versus 4.0 and 2.6 versus 4.6 respectively; both P < 0.001). The iDAScore for 0-10% fragmentation was significantly higher compared with the groups with higher fragmentation (P < 0.001). When combining cell numbers and fragmentation, iDAScore values decreased as fragmentation increased, regardless of cell number. iDAScore discriminated between embryos that resulted in live birth or no live birth (AUC of 0.627 and 0.607), compared with the morphological model (AUC of 0.618 and 0.585) for day 2 and day 3, respectively. CONCLUSIONS: The iDAScore v2.0 values correlated significantly with cell numbers and fragmentation scored manually for cleavage-stage embryos on days 2 and 3. iDAScore had some predictive value for live birth, conditional that embryo selection was based on morphology.

Discard or not discard, that is the question: an international survey across 117 embryologists on the clinical management of borderline quality blastocysts.

STUDY QUESTION: Do embryologists from different European countries agree on embryo disposition decisions ('use' or 'discard') about Day 7 (>144 h post-insemination) and/or low-quality blastocysts (LQB;

Towards Automation in IVF: Pre-Clinical Validation of a Deep Learning-Based Embryo Grading System during PGT-A Cycles.

Preimplantation genetic testing for aneuploidies (PGT-A) is arguably the most effective embryo selection strategy. Nevertheless, it requires greater workload, costs, and expertise. Therefore, a quest towards user-friendly, non-invasive strategies is ongoing. Although insufficient to replace PGT-A, embryo morphological evaluation is significantly associated with embryonic competence, but scarcely reproducible. Recently, artificial intelligence-powered analyses have been proposed to objectify and automate image evaluations. iDAScore v1.0 is a deep-learning model based on a 3D convolutional neural network trained on time-lapse videos from implanted and non-implanted blastocysts. It is a decision support system for ranking blastocysts without manual input. This retrospective, pre-clinical, external validation included 3604 blastocysts and 808 euploid transfers from 1232 cycles. All blastocysts were retrospectively assessed through the iDAScore v1.0; therefore, it did not influence embryologists' decision-making process. iDAScore v1.0 was significantly associated with embryo morphology and competence, although AUCs for euploidy and live-birth prediction were 0.60 and 0.66, respectively, which is rather comparable to embryologists' performance. Nevertheless, iDAScore v1.0 is objective and reproducible, while embryologists' evaluations are not. In a retrospective simulation, iDAScore v1.0 would have ranked euploid blastocysts as top quality in 63% of cases with one or more euploid and aneuploid blastocysts, and it would have questioned embryologists' ranking in 48% of cases with two or more euploid blastocysts and one or more live birth. Therefore, iDAScore v1.0 may objectify embryologists' evaluations, but randomized controlled trials are required to assess its clinical value.

A comparison between the Felix™ electrophoretic system of sperm isolation and conventional density gradient centrifugation: a multicentre analysis.

PURPOSE: Developing optimized techniques for the isolation of human spermatozoa possessing low levels of DNA damage is an important objective for the ART industry. The purpose of this study was to compare a novel electrophoretic system (Felix™) of sperm isolation with a conventional method involving density gradient centrifugation (DGC). METHODS: Five international ART Centres in Australia, India, Sweden, the USA, and China have collaborated in order to compare the quality of the sperm populations isolated by Felix™ and DGC in terms of processing time, sperm concentration, motility, vitality, and DNA integrity as assessed by 3 methods: SCSA, Halo, and TUNEL. RESULTS: Across all centers, 112 comparisons were performed. Although significant differences were noted between centers in terms of the quality of the semen samples subjected for analysis, overall, both methods were equally capable of isolating populations of spermatozoa exhibiting high levels of vitality and progressive motility. The absolute numbers of spermatozoa recovered were significantly (p < 0.001) lower with the Felix™ device although sperm quality was higher with 4/5 centers reporting a significant improvement in DNA integrity relative to DGC (p < 0.01-p < 0.001). In practical terms, the Felix™ device featured a standardized 6 min preparation time whereas clinical DGC protocols varied from center to center but generally took around 40 min to complete. CONCLUSIONS: The Felix™ device is a positive technical development capable of isolating suspensions of highly motile spermatozoa exhibiting low levels of DNA damage in a fraction of the time taken by conventional procedures such as DGC.

A double-blind randomized controlled trial investigating a time-lapse algorithm for selecting Day 5 blastocysts for transfer.

STUDY QUESTION: Can use of a commercially available time-lapse algorithm for Day 5 blastocyst selection improve pregnancy rates compared with morphology alone? SUMMARY ANSWER: The use of a time-lapse selection model to choose blastocysts for fresh single embryo transfer on Day 5 did not improve ongoing pregnancy rate compared to morphology alone. WHAT IS KNOWN ALREADY: Evidence from time-lapse monitoring suggests correlations between timing of key developmental events and embryo viability. No good quality evidence exists to support improved pregnancy rates following time-lapse selection. STUDY DESIGN, SIZE, DURATION: A prospective multicenter randomized controlled trial including 776 randomized patients was performed between 2018 and 2021. Patients with at least two good quality blastocysts on Day 5 were allocated by a computer randomization program in a proportion of 1:1 into either the control group, whereby single blastocysts were selected for transfer by morphology alone, or the intervention group whereby final selection was decided by a commercially available time-lapse model. The embryologists at the time of blastocyst morphological scoring were blinded to which study group the patients would be randomized, and the physician and patients were blind to which group they were allocated until after the primary outcome was known. The primary outcome was number of ongoing pregnancies in the two groups. PARTICIPANTS/MATERIALS, SETTING, METHODS: From 10 Nordic IVF clinics, 776 patients with a minimum of two good quality blastocysts on Day 5 (D5) were randomized into one of the two study groups. A commercial time-lapse model decided the final selection of blastocysts for 387 patients in the intervention (time-lapse) group, and blastocysts with the highest morphological score were transferred for 389 patients in the control group. Only single embryo transfers in fresh cycles were performed. MAIN RESULTS AND THE ROLE OF CHANCE: In the full analysis set, the ongoing pregnancy rate for the time-lapse group was 47.4% (175/369) and 48.1% (181/376) in the control group. No statistically significant difference was found between the two groups: mean difference -0.7% (95% CI -8.2, 6.7, P = 0.90). Pregnancy rate (60.2% versus 59.0%, mean difference 1.1%, 95% CI -6.2, 8.4, P = 0.81) and early pregnancy loss (21.2% versus 18.5%, mean difference 2.7%, 95% CI -5.2, 10.6, P = 0.55) were the same for the time-lapse and the control group. Subgroup analyses showed that patient and treatment characteristics did not significantly affect the commercial time-lapse model D5 performance. In the time-lapse group, the choice of best blastocyst changed on 42% of occasions (154/369, 95% CI 36.9, 47.2) after the algorithm was applied, and this rate was similar for most treatment clinics. LIMITATIONS, REASONS FOR CAUTION: During 2020, the patient recruitment rate slowed down at participating clinics owing to coronavirus disease-19 restrictions, so the target sample size was not achieved as planned and it was decided to stop the trial prematurely. The study only investigated embryo selection at the blastocyst stage on D5 in fresh IVF transfer cycles. In addition, only blastocysts of good morphological quality were considered for transfer, limiting the number of embryos for selection in both groups: also, it could be argued that this manual preselection of blastocysts limits the theoretical selection power of time-lapse, as well as restricting the results mainly to a good prognosis patient group. Most patients were aimed for blastocyst stage transfer when a minimum of five zygotes were available for extended culture. Finally, the primary clinical outcome evaluated was pregnancy to only 6-8 weeks. WIDER IMPLICATIONS OF THE FINDINGS: The study suggests that time-lapse selection with a commercially available time-lapse model does not increase chance of ongoing pregnancy after single blastocyst transfer on Day 5 compared to morphology alone. STUDY FUNDING/COMPETING INTEREST(S): The study was financed by a grant from the Swedish state under the ALF-agreement between the Swedish government and the county councils (ALFGBG-723141). Vitrolife supported the study with embryo culture dishes and culture media. During the study period, T.H. changed his employment from Livio AB to Vitrolife AB. All other authors have no conflicts of interests to disclose. TRIAL REGISTRATION NUMBER: ClinicalTrials.gov registration number NCT03445923. TRIAL REGISTRATION DATE: 26 February 2018. DATE OF FIRST PATIENT'S ENROLMENT: 11 June 2018.

Comprehensive analysis of soluble RNAs in human embryo culture media and blastocoel fluid.

PURPOSE: miRNAs have been suggested as biomarkers of embryo viability; however, findings from preliminary studies are divergent. Furthermore, the presence of other types of small RNA molecules remains to be investigated. The purpose of this study was to perform a comprehensive analysis of small non-coding RNA levels in spent and unconditioned embryo culture media, along with miRNA levels in blastocoelic fluid samples from human embryos. METHODS: miRNAs in unconditioned culture medium from 3 different manufacturers, along with miRNA from day 5 conditioned culture medium, control medium, and corresponding blastocoel fluid from 10 human blastocysts were analyzed with array-based q-PCR analysis. Subsequently, deep sequencing of total and small RNA in day 5 spent culture medium from 5 human blastocysts and corresponding controls was performed. RESULTS: In spite of using state-of-the-art sensitive detection methods, no miRNAs were found to be reliably present in the spent culture medium or the blastocoel fluid. Ct values were above the recommended limit for detection in the array-based analysis, a finding that was confirmed by deep sequencing. The majority of miRNAs identified by deep sequencing were expressed in all samples including control media and seem to originate from sources other than conditioned IVF media. CONCLUSIONS: Our findings question the use of miRNAs as a reliable biomarker and highlight the need for a critical methodological approach in miRNA studies. Interestingly, tiRNA fragments appear to be overexpressed in conditioned IVF media samples and could potentially be a novel biomarker worthy of investigation.

Conventional morphology performs better than morphokinetics for prediction of live birth after day 2 transfer.

Numerous studies have reported on the potential value of time-lapse variables for prediction of embryo viability. However, these variables have not been evaluated in combination with conventional morphological grading and patient characteristics. The aim of this study was to assess the ability of patient characteristics and embryo morphology together with morphokinetic variables to predict live birth after day 2 transfer. This retrospective analysis included 207 transferred embryos from 199 couples cultured in a time-lapse system up to day 2 of development. Good prediction of live birth or ranking of embryos with respect to live birth potential was achieved with early cleavage combined with fragmentation grade at 43-45 h. These variables were selected as the strongest predictors of live birth, as assessed by stepwise logistic regression, and additional inclusion of morphokinetic variables did not improve the model significantly. Also, neither logistic regression models nor classification tree models with morphokinetic variables were able to achieve equally good prediction of live birth, as measured by AUC on an external data set not used for model development. In conclusion, for fresh day 2 transfers early cleavage in combination with fragmentation grade at 43-45 h should be considered when selecting between good quality embryos.

Quality control and standardization of embryo morphology scoring and viability markers.

A so-called 'good-quality embryo' may be defined as an embryo that has the potential to implant into the uterine endometrium and give rise to the birth of a healthy child. A standardized and objective scoring of embryo 'quality' is therefore crucial in the classification and selection of embryos. However, embryo scoring is still being performed mainly via ocular evaluation, which often results in different interpretations of embryo quality. The addition of viability markers, such as measuring gene expression or the uptake/release of metabolites, proteins or RNA/DNA molecules in the culture media, would increase the possibility of standardized measurements. However, no single biomarker has yet been introduced into standard clinical practice, mainly due to the complexity of the techniques and the influence of biological variations and differences in culture conditions. In this paper different methods for the scoring of embryos and the possibility of standardizing and implementing quality control systems are discussed.

Cross-validation and predictive value of near-infrared spectroscopy algorithms for day-5 blastocyst transfer.

Near-infrared (NIR) spectroscopic metabolomic profiling of spent embryo-culture media has been used to calculate a viability score for individual embryos. These scores have been found to correlate to the reproductive potential of cleavage-stage embryos. In this study, 137 spent blastocyst media samples were collected after single-embryo transfer and analysed by NIR spectroscopy to generate an algorithm and calculate viability scores. To blindly validate the algorithm development process, another algorithm was trained on 47 preselected samples from clinic 1 and then used to predict the outcome of 42 samples from clinic 2. The overall pregnancy rate from the two clinical sites was 50.4%. A positive correlation (R(2)=0.82, P=0.03) was observed with the increasing viability score quintiles and their associated implantation rates. Cross-validation of an algorithm generated from NIR analysis of media samples at one clinical setting blindly was shown to predict implantation potential of blastocysts cultured at another clinic in a different culture media and culture volume. This study demonstrates that metabolomic profiling by NIR spectroscopic analysis of day-5 spent embryo-culture media can predict the implantation potential of blastocysts. Furthermore, this method may not be restricted to a specific set of culturing conditions. The successes of IVF treatment cycles are in part limited by the ability to select the best single embryo from a cohort of patient embryos for transfer back to the woman. Routine procedures of embryo selection are based on morphology, including cell number and size, and the timing of cell division. These methods are favoured because they are quick and easy to assess. Human embryos are grown in culture solutions, which are specific for their stage of development. Recent studies analysing the culture solution in which the embryo are grown, by near infrared (NIR) spectroscopic analysis, have been able to predict if an embryo will implant or not. As culture conditions often vary between IVF laboratories the questions remained if the NIR technique could be used to independently predict the implantation potential of an embryo cultured at one laboratory using an algorithm trained on embryos at a second clinic, a so-called cross-validation. The results of this study show that NIR spectroscopy can predict the ability of embryos to implant even when grown in different IVF laboratories and in two different culture solutions. This information supports the idea that NIR spectroscopy can be used globally not relying on specific culture conditions or media.

Seminal fluid drives expansion of the CD4+CD25+ T regulatory cell pool and induces tolerance to paternal alloantigens in mice.

T regulatory (Treg) cells are implicated in maternal immune tolerance of the conceptus at implantation; however, the antigenic and regulatory signals controlling Treg cells in early pregnancy are undefined. To examine the role of male seminal fluid in tolerance induction, the effect of exposure to seminal fluid at mating on responsiveness to paternal alloantigens was examined using paternal tumor cell grafts and by delayed-type hypersensitivity (DTH) challenge on Day 3.5 postcoitum. Exposure to seminal fluid inhibited rejection of paternal tumor cells, independently of fertilization and embryo development, while seminal fluid from major histocompatability complex (MHC)-dissimilar males was less effective. Similarly, mating with intact males suppressed the DTH response to paternal alloantigens in an MHC-specific fashion. Excision of the seminal vesicle glands diminished the tolerance-inducing activity of seminal fluid. Mating with intact males caused an increase in CD4(+)CD25(+) cells expressing FOXP3 in the para-aortic lymph nodes draining the uterus, beyond the estrus-associated peak in cycling mice. The increase in CD4(+)CD25(+) cells was abrogated when males were vasectomized or seminal vesicles were excised. Collectively, these data provide evidence that exposure to seminal fluid at mating promotes a state of functional tolerance to paternal alloantigens that may facilitate maternal acceptance of the conceptus at implantation, and the effects of seminal fluid are likely to be mediated by expansion of the Treg cell pool. Both seminal plasma and sperm components of the seminal fluid are necessary to confer full tolerance and elicit the Treg cell response, potentially through provision of immune-deviating cytokines and antigens, respectively.