A novel deep learning approach to identify embryo morphokinetics in multiple time lapse systems

Abstract The use of time lapse systems (TLS) in In Vitro Fertilization (IVF) labs to record developing embryos has paved the way for deep-learning based computer vision algorithms to assist embryologists in their morphokinetic evaluation. Today, most of the literature has characterized algorithms th...

Full description

Saved in:
Bibliographic Details
Main Authors: Guillaume Canat, Antonin Duval, Nina Gidel-Dissler, Alexandra Boussommier-Calleja
Format: Article
Language:English
Published: Nature Portfolio 2024-11-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-024-80565-1
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850163295740231680
author Guillaume Canat
Antonin Duval
Nina Gidel-Dissler
Alexandra Boussommier-Calleja
author_facet Guillaume Canat
Antonin Duval
Nina Gidel-Dissler
Alexandra Boussommier-Calleja
author_sort Guillaume Canat
collection DOAJ
description Abstract The use of time lapse systems (TLS) in In Vitro Fertilization (IVF) labs to record developing embryos has paved the way for deep-learning based computer vision algorithms to assist embryologists in their morphokinetic evaluation. Today, most of the literature has characterized algorithms that predict pregnancy, ploidy or blastocyst quality, leaving to the side the task of identifying key morphokinetic events. Using a dataset of N = 1909 embryos collected from multiple clinics equipped with EMBRYOSCOPE/EMBRYOSCOPE+ (Vitrolife), GERI (Genea Biomedx) or MIRI (Esco Medical), this study proposes a novel deep-learning architecture to automatically detect 11 kinetic events (from 1-cell to blastocyst). First, a Transformer based video backbone was trained with a custom metric inspired by reverse cross-entropy which enables the model to learn the ordinal structure of the events. Second, embeddings were extracted from the backbone and passed into a Gated Recurrent Unit (GRU) sequence model to account for kinetic dependencies. A weighted average of 66.0%, 67.6% and 66.3% in timing precision, recall and F1-score respectively was reached on a test set of 278 embryos, with a model applicable to multiple TLS.
format Article
id doaj-art-a6570c06d27944b0b79a5bc0bcdf474f
institution OA Journals
issn 2045-2322
language English
publishDate 2024-11-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-a6570c06d27944b0b79a5bc0bcdf474f2025-08-20T02:22:20ZengNature PortfolioScientific Reports2045-23222024-11-011411910.1038/s41598-024-80565-1A novel deep learning approach to identify embryo morphokinetics in multiple time lapse systemsGuillaume Canat0Antonin Duval1Nina Gidel-Dissler2Alexandra Boussommier-Calleja3ImVitro, AI TeamImVitro, AI TeamImVitro, AI TeamImVitro, AI TeamAbstract The use of time lapse systems (TLS) in In Vitro Fertilization (IVF) labs to record developing embryos has paved the way for deep-learning based computer vision algorithms to assist embryologists in their morphokinetic evaluation. Today, most of the literature has characterized algorithms that predict pregnancy, ploidy or blastocyst quality, leaving to the side the task of identifying key morphokinetic events. Using a dataset of N = 1909 embryos collected from multiple clinics equipped with EMBRYOSCOPE/EMBRYOSCOPE+ (Vitrolife), GERI (Genea Biomedx) or MIRI (Esco Medical), this study proposes a novel deep-learning architecture to automatically detect 11 kinetic events (from 1-cell to blastocyst). First, a Transformer based video backbone was trained with a custom metric inspired by reverse cross-entropy which enables the model to learn the ordinal structure of the events. Second, embeddings were extracted from the backbone and passed into a Gated Recurrent Unit (GRU) sequence model to account for kinetic dependencies. A weighted average of 66.0%, 67.6% and 66.3% in timing precision, recall and F1-score respectively was reached on a test set of 278 embryos, with a model applicable to multiple TLS.https://doi.org/10.1038/s41598-024-80565-1EmbryologyMachine learningMedical Imaging
spellingShingle Guillaume Canat
Antonin Duval
Nina Gidel-Dissler
Alexandra Boussommier-Calleja
A novel deep learning approach to identify embryo morphokinetics in multiple time lapse systems
Scientific Reports
Embryology
Machine learning
Medical Imaging
title A novel deep learning approach to identify embryo morphokinetics in multiple time lapse systems
title_full A novel deep learning approach to identify embryo morphokinetics in multiple time lapse systems
title_fullStr A novel deep learning approach to identify embryo morphokinetics in multiple time lapse systems
title_full_unstemmed A novel deep learning approach to identify embryo morphokinetics in multiple time lapse systems
title_short A novel deep learning approach to identify embryo morphokinetics in multiple time lapse systems
title_sort novel deep learning approach to identify embryo morphokinetics in multiple time lapse systems
topic Embryology
Machine learning
Medical Imaging
url https://doi.org/10.1038/s41598-024-80565-1
work_keys_str_mv AT guillaumecanat anoveldeeplearningapproachtoidentifyembryomorphokineticsinmultipletimelapsesystems
AT antoninduval anoveldeeplearningapproachtoidentifyembryomorphokineticsinmultipletimelapsesystems
AT ninagideldissler anoveldeeplearningapproachtoidentifyembryomorphokineticsinmultipletimelapsesystems
AT alexandraboussommiercalleja anoveldeeplearningapproachtoidentifyembryomorphokineticsinmultipletimelapsesystems
AT guillaumecanat noveldeeplearningapproachtoidentifyembryomorphokineticsinmultipletimelapsesystems
AT antoninduval noveldeeplearningapproachtoidentifyembryomorphokineticsinmultipletimelapsesystems
AT ninagideldissler noveldeeplearningapproachtoidentifyembryomorphokineticsinmultipletimelapsesystems
AT alexandraboussommiercalleja noveldeeplearningapproachtoidentifyembryomorphokineticsinmultipletimelapsesystems