Benchmarking multiple instance learning architectures from patches to pathology for prostate cancer detection and grading using attention-based weak supervision

Butt, Naveed Anwer and Sarwat, Dilawaiz and Delgado Noya, Irene and Tutusaus, Kilian and Samee, Nagwan Abdel and Ashraf, Imran UNSPECIFIED, UNSPECIFIED, irene.delgado@uneatlantico.es, kilian.tutusaus@uneatlantico.es, UNSPECIFIED, UNSPECIFIED (2026) Benchmarking multiple instance learning architectures from patches to pathology for prostate cancer detection and grading using attention-based weak supervision. Scientific Reports. ISSN 2045-2322

Text s41598-026-39196-x_reference.pdf Download (2MB)

Abstract

Histopathological evaluation is necessary for the diagnosis and grading of prostate cancer, which is still one of the most common cancers in men globally. Traditional evaluation is time-consuming, prone to inter-observer variability, and challenging to scale. The clinical usefulness of current AI systems is limited by the need for comprehensive pixel-level annotations. The objective of this research is to develop and evaluate a large-scale benchmarking study on a weakly supervised deep learning framework that minimizes the need for annotation and ensures interpretability for automated prostate cancer diagnosis and International Society of Urological Pathology (ISUP) grading using whole slide images (WSIs). This study rigorously tested six cutting-edge multiple instance learning (MIL) architectures (CLAM-MB, CLAM-SB, ILRA-MIL, AC-MIL, AMD-MIL, WiKG-MIL), three feature encoders (ResNet50, CTransPath, UNI2), and four patch extraction techniques (varying sizes and overlap) using the PANDA dataset (10,616 WSIs), yielding 72 experimental configurations. The methodology used distributed cloud computing to process over 31 million tissue patches, implementing advanced attention mechanisms to ensure clinical interpretability through Grad-CAM visualizations. The optimum configuration (UNI2 encoder with ILRA-MIL, 256 256 patches, 50% overlap) achieved 78.75% accuracy and 90.12% quadratic weighted kappa (QWK), outperforming traditional methods and approaching expert pathologist-level diagnostic capability. Overlapping smaller patches offered the best balance of spatial resolution and contextual information, while domain-specific foundation models performed noticeably better than generic encoders. This work is the first large-scale, comprehensive comparison of weekly supervised MIL methods for prostate cancer diagnosis and grading. The proposed approach has excellent clinical diagnostic performance, scalability, practical feasibility through cloud computing, and interpretability using visualization tools.

Item Type:	Article
Uncontrolled Keywords:	Prostate cancer detection Weakly supervised learning Multiple instance learning Whole slide images ISUP grading
Subjects:	Subjects > Biomedicine Subjects > Engineering
Divisions:	Europe University of Atlantic > Research > Scientific Production Fundación Universitaria Internacional de Colombia > Research > Scientific Production Ibero-american International University > Research > Scientific Production Ibero-american International University > Research > Scientific Production Universidad Internacional do Cuanza > Research > Scientific Production University of La Romana > Research > Scientific Production
Depositing User:	Sr Bibliotecario
Date Deposited:	13 Mar 2026 12:35
Last Modified:	13 Mar 2026 12:35
URI:	http://repositorio.funiber.org/id/eprint/27825

Actions (login required)

View Item