LWMA-Net: Light-weighted morphology attention learning for human embryo grading

Visual inspection of embryo morphology is routinely used in embryo assessment and selection. However, due to the complexity of morphologies and large inter- and intra-observer variances among embryologists, manual evaluations remain to be subjective and time-consuming. Thus, we proposed a light-weighted morphology attention learning network (LWMA-Net) for automatic assistance on embryo grading. The LWMA-Net integrated a morphology attention module (MAM) to seek the informative features and their locations and a multiscale fusion module (MFM) to increase the features flowing in the model. The LWMA-Net was trained with a primary set of 3599 embryos from 2318 couples that were clinically enrolled between Sep. 2016 and Dec. 2018, and generated area under the receiver operating characteristic curves (AUCs) of 96.88% and 97.58% on 4- and 3-category gradings, respectively. An independent test set comprises 691 embryos from 321 couples between Jan. 2019 and Jan. 2021 were used to test the assisted fertility values on the embryo grading. Five experienced embryologists were invited to regrade the embryos in the independent set with and without the aid of the LWMA-Net three months apart. Embryologists aided by our LWMA-Net significantly improved their grading capabilities with average AUCs improved by 4.98%-5.32% on 4- and 3-category grading tasks, respectively, which suggests good potential of our LWMA-Net on assisted human reproduction.

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

BACKGROUND

Embryo morphology is a predictive marker for implantation success and ultimately live births. Viability evaluation and quality grading are commonly used to select the embryo with the highest implantation potential. However, the traditional method of manual embryo assessment is time-consuming and highly susceptible to inter- and intra-observer variability. Automation of this process results in more objective and accurate predictions.

METHOD

In this paper, we propose a novel methodology based on deep learning to automatically evaluate the morphological appearance of human embryos from time-lapse imaging. A supervised contrastive learning framework is implemented to predict embryo viability at day 4 and day 5, and an inductive transfer approach is applied to classify embryo quality at both times.

RESULTS

Results showed that both methods outperformed conventional approaches and improved state-of-the-art embryology results for an independent test set. The viability result achieved an accuracy of 0.8103 and 0.9330 and the quality results reached values of 0.7500 and 0.8001 for day 4 and day 5, respectively. Furthermore, qualitative results kept consistency with the clinical interpretation.

CONCLUSIONS

The proposed methods are up to date with the artificial intelligence literature and have been proven to be promising. Furthermore, our findings represent a breakthrough in the field of embryology in that they study the possibilities of embryo selection at day 4. Moreover, the grad-CAMs findings are directly in line with embryologists' decisions. Finally, our results demonstrated excellent potential for the inclusion of the models in clinical practice.

Inter-centre reliability in embryo grading across several IVF clinics is limited: implications for embryo selection

RESEARCH QUESTION

What is the intra- and inter-centre reliability in embryo grading performed according to the Istanbul Consensus across several IVF clinics?

DESIGN

Forty Day 3 embryos and 40 blastocysts were photographed on three focal planes. Senior and junior embryologists from 65 clinics were invited to grade them according to the Istanbul Consensus (Study Phase I). All participants then attended interactive training where a panel of experts graded the same embryos (Study Phase II). Finally, a second set of pictures was sent to both embryologists and experts for a blinded evaluation (Study Phase III). Intra-centre reliability was reported for Study Phase I as Cohen's kappa between senior and junior embryologists; inter-centre reliability was instead calculated between senior/junior embryologists and experts in Study Phase I versus III to outline improvements after training (i.e. upgrade of Cohen's kappa category according to Landis and Koch).

RESULTS

Thirty-six embryologists from 18 centres participated (28% participation rate). The intra-centre reliability was (i) substantial (0.63) for blastomere symmetry (range -0.02 to 1.0), (ii) substantial (0.72) for fragmentation (range 0.29-1.0), (iii) substantial (0.66) for blastocyst expansion (range 0.19-1.0), (iv) moderate (0.59) for inner cell mass quality (range 0.07-0.92), (v) moderate (0.56) for trophectoderm quality (range 0.01-0.97). The inter-centre reliability showed an overall improvement from Study Phase I to III, from fair (0.21-0.4) to moderate (0.41-0.6) for all parameters under analysis, except for blastomere fragmentation among senior embryologists, which was already moderate before training.

CONCLUSIONS

Intra-centre reliability was generally moderate/substantial, while inter-centre reliability was just fair. The interactive training improved it to moderate, hence this workflow was deemed helpful. The establishment of external quality assessment services (e.g. UK NEQAS) and the avant-garde of artificial intelligence might further improve the reliability of this key practice for embryo selection.