Interactive Multiple Instance Learning Network for Whole Slide Image Analysis

doi:10.1016/j.eswa.2025.129338

Interactive Multiple Instance Learning Network for Whole Slide Image Analysis

2025-08-21

0

PRE

AI

DOI：10.1016/j.eswa.2025.129338

Original

Original request for help

Save

Abstract

En 中文

• Introduces iMIL (Interactive Multiple Instance Learning) for Whole Slide Image (WSI) classification, which can integrates both visual patterns and clinical textual information • iMILcan better handle of weak supervision scenarios, improve ability to identify diagnostically crucial regions, and effective integrate of multi-modal information. • This article presents a significant advancement in WSI analysis by combining multiple innovative approaches and demonstrating clear performance improvements over existing methods.

Keywords:

iMIL

Whole Slide Image

weak supervision

multi-modal integration

diagnostic regions

Interactive Multiple Instance Learning Network for Whole Slide Image Analysis

Abstract

Journal

Organization