From Division to Decision: Leveraging Temporal Cell-Stage Segmentation for Embryo Transferability Prediction

TL;DR

This paper introduces TransFACT, a transformer-based model that uses early developmental stage segmentation from time-lapse videos to improve bovine embryo transferability prediction, surpassing existing methods.

Contribution

The novel TransFACT framework leverages stage-level supervision and action recognition techniques for better embryo transferability prediction from early videos.

Findings

TransFACT outperforms existing methods in transferability prediction.

Stage-level representations improve prediction accuracy.

Using developmental stages as auxiliary supervision enhances model performance.

Abstract

Accurate selection of bovine embryos is a challenging task, as current practice relies on a single expert assessment on the seventh day after insemination, resulting in high rates of pregnancy loss. Time-lapse videomicroscopy provides detailed information on early development, but is difficult to exploit because of complex motion patterns and time-consuming analysis. We propose TransFACT, a transformer-based framework for modeling early developmental stages and embryo transferability using 2D time-lapse videos from the first four days of development. TransFACT combines frame-level temporal features with stage-level representations, using developmental stages as auxiliary supervision to predict transferability on day four. Our experiments demonstrate that TransFACT, by leveraging an existing method designed for action recognition, achieves superior performance than its competitor in predicting embryo transferability.