Attention-Based Whole-Slide Image Compression Achieves Pathologist-Level Prescreening of Multiorgan Routine Histopathology Biopsies
Modern Pathology · 2025-11 · 38 (11) · 100827
DOI: 10.1016/j.modpat.2025.100827 ↗Kurz gefasst
Fachartikel in der Zeitschrift Modern Pathology, 2025-11. Verfasst von Witali Aswolinskiy und 10 weiteren Autor:innen. Beteiligte Einrichtung: Department of Pathology, Radboud University Medical Center, Nijmegen, The Netherlands. Hauptthemen laut MEDLINE: Deep Learning, Data Compression, Image Interpretation, Computer-Assisted, Early Detection of Cancer. Die Arbeit wurde bisher 8-mal zitiert (Crossref) und nennt 26 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Horizon 2020 Framework Programme.
Abstract (Originalsprache)
Screening programs for the early detection of cancers, such as colorectal and cervical cancers, have led to an increased demand for histopathological analysis of biopsies. Advanced image analysis with deep learning has shown the potential to automate cancer detection in digital pathology whole-slide images. In particular, weakly supervised learning can achieve whole-slide image classification without the need for tedious, manual annotations, using only slide-level labels. Here, we used data from n = 12,580 whole-slide images from n = 9141 tissue blocks to train and validate a weakly supervised deep learning approach based on Neural Image Compression with Attention (NIC-A) using labels extracted from pathology reports. We also introduced slide packing, a method that merges tissue from multiple slides of the same tissue block into a single "packed" image linked to block-level labels. NIC-A classifies colon and cervical tissue slides into cancer, high-grade dysplasia, low-grade dysplasia, and normal tissue and detects celiac disease in duodenal biopsies. We validated NIC-A for colon and cervix against panels of 4 and 3 pathologists, respectively, on cohorts from 2 European centers. We show that the proposed approach reaches pathologist-level performance in detecting and classifying abnormalities, suggesting its potential to assist pathologists in prescreening workflows by reducing workload in routine digital pathology diagnostics.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Pathology, Radboud University Medical Center, Nijmegen, The Netherlands.
- Department of Pathology, Santa Chiara Hospital, Trento, Italy.
- Department of Pathology, ASST Spedali Civili, Brescia, Italy.
- Pathology Unit, Gravina Hospital Caltagirone ASP, Catania, Italy.
- Department of Pathology, Radboud University Medical Center, Nijmegen, The Netherlands; Center for Medical Image Science and Visualization, Linköping University, Linköping, Sweden.
- Department of Pathology and Diagnostics, University and Hospital Trust of Verona, Verona, Italy.
Förderung
- Horizon 2020 Framework Programme
Zitieren
APA
Aswolinskiy, W., van der Post, R. S., Campora, M., Baronchelli, C., Ardighieri, L., Vatrano, S., van der Laak, J., Munari, E., Simons, M., Nagtegaal, I., & Ciompi, F. (2025). Attention-Based Whole-Slide Image Compression Achieves Pathologist-Level Prescreening of Multiorgan Routine Histopathology Biopsies. Modern Pathology, 38(11), 100827. https://doi.org/10.1016/j.modpat.2025.100827
Vancouver
Aswolinskiy W, van der Post RS, Campora M, Baronchelli C, Ardighieri L, Vatrano S, et al. Attention-Based Whole-Slide Image Compression Achieves Pathologist-Level Prescreening of Multiorgan Routine Histopathology Biopsies. Modern Pathology. 2025;38(11):100827. doi:10.1016/j.modpat.2025.100827
BibTeX
@article{Aswolinskiy2025,
author = {Aswolinskiy, Witali and van der Post, Rachel S and Campora, Michela and Baronchelli, Carla and Ardighieri, Laura and Vatrano, Simona and van der Laak, Jeroen and Munari, Enrico and Simons, Michiel and Nagtegaal, Iris and Ciompi, Francesco},
title = {Attention-Based Whole-Slide Image Compression Achieves Pathologist-Level Prescreening of Multiorgan Routine Histopathology Biopsies},
journal = {Modern Pathology},
year = {2025},
volume = {38},
number = {11},
pages = {100827},
doi = {10.1016/j.modpat.2025.100827}
}
EUVIMED