A self-supervised learning approach for high throughput and high content cell segmentation

Abstract In principle, ML/AI-based algorithms should enable rapid and accurate cell segmentation in high-throughput settings. However, reliance on large training datasets, human input, computational expertise, and limited generalizability has prevented this goal of completely automated, high-through...

Full description

Saved in:
Bibliographic Details
Main Authors: Van K. Lam, Jeff M. Byers, Michael C. Robitaille, Logan Kaler, Joseph A. Christodoulides, Marc P. Raphael
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:Communications Biology
Online Access:https://doi.org/10.1038/s42003-025-08190-w
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849325988995923968
author Van K. Lam
Jeff M. Byers
Michael C. Robitaille
Logan Kaler
Joseph A. Christodoulides
Marc P. Raphael
author_facet Van K. Lam
Jeff M. Byers
Michael C. Robitaille
Logan Kaler
Joseph A. Christodoulides
Marc P. Raphael
author_sort Van K. Lam
collection DOAJ
description Abstract In principle, ML/AI-based algorithms should enable rapid and accurate cell segmentation in high-throughput settings. However, reliance on large training datasets, human input, computational expertise, and limited generalizability has prevented this goal of completely automated, high-throughput segmentation from being achieved. To overcome these roadblocks, we introduce an innovative self-supervised learning method (SSL) for pixel classification that does not require parameter tuning or curated data sets, and instead trains itself on the end-users’ own data in a completely automated fashion, thus providing a more efficient cell segmentation approach for high-throughput, high-content image analysis. We demonstrate that our algorithm meets the criteria of being fully automated with versatility across various magnifications, optical modalities, and cell types. Moreover, our SSL algorithm is capable of identifying complex cellular structures and organelles, which are otherwise easily missed, thereby broadening the machine learning applications to high-content imaging. Our SSL technique displayed consistently high F1 scores across segmented cell images, with scores ranging from 0.771 to 0.888, matching or outperforming the popular Cellpose algorithm, which showed a greater F1 variance of 0.454 to 0.882, primarily due to more false negatives.
format Article
id doaj-art-b4092a1a5f3246d9a78ebe6bc50612da
institution Kabale University
issn 2399-3642
language English
publishDate 2025-05-01
publisher Nature Portfolio
record_format Article
series Communications Biology
spelling doaj-art-b4092a1a5f3246d9a78ebe6bc50612da2025-08-20T03:48:15ZengNature PortfolioCommunications Biology2399-36422025-05-018111010.1038/s42003-025-08190-wA self-supervised learning approach for high throughput and high content cell segmentationVan K. Lam0Jeff M. Byers1Michael C. Robitaille2Logan Kaler3Joseph A. Christodoulides4Marc P. Raphael5US Naval Research LaboratoryUS Naval Research LaboratoryUS Naval Research LaboratoryUS Naval Research LaboratoryUS Naval Research LaboratoryUS Naval Research LaboratoryAbstract In principle, ML/AI-based algorithms should enable rapid and accurate cell segmentation in high-throughput settings. However, reliance on large training datasets, human input, computational expertise, and limited generalizability has prevented this goal of completely automated, high-throughput segmentation from being achieved. To overcome these roadblocks, we introduce an innovative self-supervised learning method (SSL) for pixel classification that does not require parameter tuning or curated data sets, and instead trains itself on the end-users’ own data in a completely automated fashion, thus providing a more efficient cell segmentation approach for high-throughput, high-content image analysis. We demonstrate that our algorithm meets the criteria of being fully automated with versatility across various magnifications, optical modalities, and cell types. Moreover, our SSL algorithm is capable of identifying complex cellular structures and organelles, which are otherwise easily missed, thereby broadening the machine learning applications to high-content imaging. Our SSL technique displayed consistently high F1 scores across segmented cell images, with scores ranging from 0.771 to 0.888, matching or outperforming the popular Cellpose algorithm, which showed a greater F1 variance of 0.454 to 0.882, primarily due to more false negatives.https://doi.org/10.1038/s42003-025-08190-w
spellingShingle Van K. Lam
Jeff M. Byers
Michael C. Robitaille
Logan Kaler
Joseph A. Christodoulides
Marc P. Raphael
A self-supervised learning approach for high throughput and high content cell segmentation
Communications Biology
title A self-supervised learning approach for high throughput and high content cell segmentation
title_full A self-supervised learning approach for high throughput and high content cell segmentation
title_fullStr A self-supervised learning approach for high throughput and high content cell segmentation
title_full_unstemmed A self-supervised learning approach for high throughput and high content cell segmentation
title_short A self-supervised learning approach for high throughput and high content cell segmentation
title_sort self supervised learning approach for high throughput and high content cell segmentation
url https://doi.org/10.1038/s42003-025-08190-w
work_keys_str_mv AT vanklam aselfsupervisedlearningapproachforhighthroughputandhighcontentcellsegmentation
AT jeffmbyers aselfsupervisedlearningapproachforhighthroughputandhighcontentcellsegmentation
AT michaelcrobitaille aselfsupervisedlearningapproachforhighthroughputandhighcontentcellsegmentation
AT logankaler aselfsupervisedlearningapproachforhighthroughputandhighcontentcellsegmentation
AT josephachristodoulides aselfsupervisedlearningapproachforhighthroughputandhighcontentcellsegmentation
AT marcpraphael aselfsupervisedlearningapproachforhighthroughputandhighcontentcellsegmentation
AT vanklam selfsupervisedlearningapproachforhighthroughputandhighcontentcellsegmentation
AT jeffmbyers selfsupervisedlearningapproachforhighthroughputandhighcontentcellsegmentation
AT michaelcrobitaille selfsupervisedlearningapproachforhighthroughputandhighcontentcellsegmentation
AT logankaler selfsupervisedlearningapproachforhighthroughputandhighcontentcellsegmentation
AT josephachristodoulides selfsupervisedlearningapproachforhighthroughputandhighcontentcellsegmentation
AT marcpraphael selfsupervisedlearningapproachforhighthroughputandhighcontentcellsegmentation