Deep learning for multiclass tumor cell detection in histopathology slides of hereditary diffuse gastric cancer

Summary: Hereditary diffuse gastric cancer (HDGC) is a rare condition where early tumor detection is challenging due to diffuse infiltration and tumor heterogeneity. Accurate identification of DGC cells is essential for understanding tumor behavior. This study aimed to develop deep learning models t...

Full description

Saved in:
Bibliographic Details
Main Authors: Robin Lomans, Valentina Angerilli, Joey Spronck, Liudmila L. Kodach, Irene Gullo, Fátima Carneiro, Rachel S. van der Post, Francesco Ciompi
Format: Article
Language:English
Published: Elsevier 2025-08-01
Series:iScience
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2589004225013252
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849716661416886272
author Robin Lomans
Valentina Angerilli
Joey Spronck
Liudmila L. Kodach
Irene Gullo
Fátima Carneiro
Rachel S. van der Post
Francesco Ciompi
author_facet Robin Lomans
Valentina Angerilli
Joey Spronck
Liudmila L. Kodach
Irene Gullo
Fátima Carneiro
Rachel S. van der Post
Francesco Ciompi
author_sort Robin Lomans
collection DOAJ
description Summary: Hereditary diffuse gastric cancer (HDGC) is a rare condition where early tumor detection is challenging due to diffuse infiltration and tumor heterogeneity. Accurate identification of DGC cells is essential for understanding tumor behavior. This study aimed to develop deep learning models to automatically detect key tumor cell types—typical and atypical signet ring cells and non-signet ring tumor cells—in H&E-stained digital pathology slides from HDGC patients. Using a multi-center dataset of 350 whole-slide images and over 91,000 annotated cells from 43 patients, we trained nnU-Net models for cell detection and compared them to Faster R-CNN baselines. We also conducted a reader study with five pathologists to benchmark performance. nnU-Net outperformed both pathologist inter-observer agreement and Faster R-CNN, achieving an F1 score of 0.49. It also matched human-level performance in estimating lesion size and cell type distributions, demonstrating its potential to support DGC tumor detection and analysis.
format Article
id doaj-art-de4a42d4edcb4089924057b167d1d561
institution DOAJ
issn 2589-0042
language English
publishDate 2025-08-01
publisher Elsevier
record_format Article
series iScience
spelling doaj-art-de4a42d4edcb4089924057b167d1d5612025-08-20T03:12:56ZengElsevieriScience2589-00422025-08-0128811306410.1016/j.isci.2025.113064Deep learning for multiclass tumor cell detection in histopathology slides of hereditary diffuse gastric cancerRobin Lomans0Valentina Angerilli1Joey Spronck2Liudmila L. Kodach3Irene Gullo4Fátima Carneiro5Rachel S. van der Post6Francesco Ciompi7Department of Pathology, Radboud University Medical Centre, Nijmegen, the Netherlands; Corresponding authorDepartment of Pathology, Radboud University Medical Centre, Nijmegen, the NetherlandsDepartment of Pathology, Radboud University Medical Centre, Nijmegen, the NetherlandsDepartment of Pathology, The Netherlands Cancer Institute, Amsterdam, the NetherlandsDepartment of Pathology, Unidade Local de Saúde São João, Porto, Portugal; Department of Pathology, Faculty of Medicine of the University of Porto (FMUP), Porto, Portugal; Instituto de Investigação E Inovação Em Saúde (i3S) & Institute of Molecular Pathology and Immunology of the University of Porto (Ipatimup), Porto, PortugalDepartment of Pathology, Unidade Local de Saúde São João, Porto, Portugal; Department of Pathology, Faculty of Medicine of the University of Porto (FMUP), Porto, Portugal; Instituto de Investigação E Inovação Em Saúde (i3S) & Institute of Molecular Pathology and Immunology of the University of Porto (Ipatimup), Porto, PortugalDepartment of Pathology, Radboud University Medical Centre, Nijmegen, the Netherlands; Corresponding authorDepartment of Pathology, Radboud University Medical Centre, Nijmegen, the Netherlands; Corresponding authorSummary: Hereditary diffuse gastric cancer (HDGC) is a rare condition where early tumor detection is challenging due to diffuse infiltration and tumor heterogeneity. Accurate identification of DGC cells is essential for understanding tumor behavior. This study aimed to develop deep learning models to automatically detect key tumor cell types—typical and atypical signet ring cells and non-signet ring tumor cells—in H&E-stained digital pathology slides from HDGC patients. Using a multi-center dataset of 350 whole-slide images and over 91,000 annotated cells from 43 patients, we trained nnU-Net models for cell detection and compared them to Faster R-CNN baselines. We also conducted a reader study with five pathologists to benchmark performance. nnU-Net outperformed both pathologist inter-observer agreement and Faster R-CNN, achieving an F1 score of 0.49. It also matched human-level performance in estimating lesion size and cell type distributions, demonstrating its potential to support DGC tumor detection and analysis.http://www.sciencedirect.com/science/article/pii/S2589004225013252CancerArtificial intelligence
spellingShingle Robin Lomans
Valentina Angerilli
Joey Spronck
Liudmila L. Kodach
Irene Gullo
Fátima Carneiro
Rachel S. van der Post
Francesco Ciompi
Deep learning for multiclass tumor cell detection in histopathology slides of hereditary diffuse gastric cancer
iScience
Cancer
Artificial intelligence
title Deep learning for multiclass tumor cell detection in histopathology slides of hereditary diffuse gastric cancer
title_full Deep learning for multiclass tumor cell detection in histopathology slides of hereditary diffuse gastric cancer
title_fullStr Deep learning for multiclass tumor cell detection in histopathology slides of hereditary diffuse gastric cancer
title_full_unstemmed Deep learning for multiclass tumor cell detection in histopathology slides of hereditary diffuse gastric cancer
title_short Deep learning for multiclass tumor cell detection in histopathology slides of hereditary diffuse gastric cancer
title_sort deep learning for multiclass tumor cell detection in histopathology slides of hereditary diffuse gastric cancer
topic Cancer
Artificial intelligence
url http://www.sciencedirect.com/science/article/pii/S2589004225013252
work_keys_str_mv AT robinlomans deeplearningformulticlasstumorcelldetectioninhistopathologyslidesofhereditarydiffusegastriccancer
AT valentinaangerilli deeplearningformulticlasstumorcelldetectioninhistopathologyslidesofhereditarydiffusegastriccancer
AT joeyspronck deeplearningformulticlasstumorcelldetectioninhistopathologyslidesofhereditarydiffusegastriccancer
AT liudmilalkodach deeplearningformulticlasstumorcelldetectioninhistopathologyslidesofhereditarydiffusegastriccancer
AT irenegullo deeplearningformulticlasstumorcelldetectioninhistopathologyslidesofhereditarydiffusegastriccancer
AT fatimacarneiro deeplearningformulticlasstumorcelldetectioninhistopathologyslidesofhereditarydiffusegastriccancer
AT rachelsvanderpost deeplearningformulticlasstumorcelldetectioninhistopathologyslidesofhereditarydiffusegastriccancer
AT francescociompi deeplearningformulticlasstumorcelldetectioninhistopathologyslidesofhereditarydiffusegastriccancer