Differential cell counts using center-point networks achieves human-level accuracy and efficiency over segmentation

Abstract Differential cell counts is a challenging task when applying computer vision algorithms to pathology. Existing approaches to train cell recognition require high availability of multi-class segmentation and/or bounding box annotations and suffer in performance when objects are tightly cluste...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Sarada M. W. Lee, Andrew Shaw, Jodie L. Simpson, David Uminsky, Luke W. Garratt
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/bc5ecf7f2ac74511b28084801ee0ef06
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:bc5ecf7f2ac74511b28084801ee0ef06
record_format	dspace
spelling	oai:doaj.org-article:bc5ecf7f2ac74511b28084801ee0ef062021-12-02T15:10:46ZDifferential cell counts using center-point networks achieves human-level accuracy and efficiency over segmentation10.1038/s41598-021-96067-32045-2322https://doaj.org/article/bc5ecf7f2ac74511b28084801ee0ef062021-08-01T00:00:00Zhttps://doi.org/10.1038/s41598-021-96067-3https://doaj.org/toc/2045-2322Abstract Differential cell counts is a challenging task when applying computer vision algorithms to pathology. Existing approaches to train cell recognition require high availability of multi-class segmentation and/or bounding box annotations and suffer in performance when objects are tightly clustered. We present differential count network (“DCNet”), an annotation efficient modality that utilises keypoint detection to locate in brightfield images the centre points of cells (not nuclei) and their cell class. The single centre point annotation for DCNet lowered burden for experts to generate ground truth data by 77.1% compared to bounding box labeling. Yet centre point annotation still enabled high accuracy when training DCNet on a multi-class algorithm on whole cell features, matching human experts in all 5 object classes in average precision and outperforming humans in consistency. The efficacy and efficiency of the DCNet end-to-end system represents a significant progress toward an open source, fully computationally approach to differential cell count based diagnosis that can be adapted to any pathology need.Sarada M. W. LeeAndrew ShawJodie L. SimpsonDavid UminskyLuke W. GarrattNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-14 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Sarada M. W. Lee Andrew Shaw Jodie L. Simpson David Uminsky Luke W. Garratt Differential cell counts using center-point networks achieves human-level accuracy and efficiency over segmentation
description	Abstract Differential cell counts is a challenging task when applying computer vision algorithms to pathology. Existing approaches to train cell recognition require high availability of multi-class segmentation and/or bounding box annotations and suffer in performance when objects are tightly clustered. We present differential count network (“DCNet”), an annotation efficient modality that utilises keypoint detection to locate in brightfield images the centre points of cells (not nuclei) and their cell class. The single centre point annotation for DCNet lowered burden for experts to generate ground truth data by 77.1% compared to bounding box labeling. Yet centre point annotation still enabled high accuracy when training DCNet on a multi-class algorithm on whole cell features, matching human experts in all 5 object classes in average precision and outperforming humans in consistency. The efficacy and efficiency of the DCNet end-to-end system represents a significant progress toward an open source, fully computationally approach to differential cell count based diagnosis that can be adapted to any pathology need.
format	article
author	Sarada M. W. Lee Andrew Shaw Jodie L. Simpson David Uminsky Luke W. Garratt
author_facet	Sarada M. W. Lee Andrew Shaw Jodie L. Simpson David Uminsky Luke W. Garratt
author_sort	Sarada M. W. Lee
title	Differential cell counts using center-point networks achieves human-level accuracy and efficiency over segmentation
title_short	Differential cell counts using center-point networks achieves human-level accuracy and efficiency over segmentation
title_full	Differential cell counts using center-point networks achieves human-level accuracy and efficiency over segmentation
title_fullStr	Differential cell counts using center-point networks achieves human-level accuracy and efficiency over segmentation
title_full_unstemmed	Differential cell counts using center-point networks achieves human-level accuracy and efficiency over segmentation
title_sort	differential cell counts using center-point networks achieves human-level accuracy and efficiency over segmentation
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/bc5ecf7f2ac74511b28084801ee0ef06
work_keys_str_mv	AT saradamwlee differentialcellcountsusingcenterpointnetworksachieveshumanlevelaccuracyandefficiencyoversegmentation AT andrewshaw differentialcellcountsusingcenterpointnetworksachieveshumanlevelaccuracyandefficiencyoversegmentation AT jodielsimpson differentialcellcountsusingcenterpointnetworksachieveshumanlevelaccuracyandefficiencyoversegmentation AT daviduminsky differentialcellcountsusingcenterpointnetworksachieveshumanlevelaccuracyandefficiencyoversegmentation AT lukewgarratt differentialcellcountsusingcenterpointnetworksachieveshumanlevelaccuracyandefficiencyoversegmentation
_version_	1718387716340580352

Differential cell counts using center-point networks achieves human-level accuracy and efficiency over segmentation

Ejemplares similares