Look a Group at Once: Multi-Slide Modeling for Survival Prediction

TL;DR

This paper introduces GroupMIL, a multi-slide modeling framework that captures cross-slide prognostic features for improved survival prediction, outperforming existing methods across multiple cancer datasets.

Contribution

The paper proposes GroupMIL and GPAMamba, novel models that jointly analyze multiple pathology slides to enhance survival prediction accuracy.

Findings

Outperforms state-of-the-art methods on five TCGA datasets.

Effectively captures cross-slide prognostic features.

Provides comprehensive survival risk assessments.

Abstract

Survival prediction is a critical task in pathology. In clinical practice, pathologists often examine multiple cases, leveraging a broader spectrum of cancer phenotypes to enhance pathological assessment. Despite significant advancements in deep learning, current solutions typically model each slide as a sample, struggling to effectively capture comparable and slide-agnostic pathological features. In this paper, we introduce GroupMIL, a novel framework inspired by the clinical practice of collective analysis, which models multiple slides as a single sample and organizes groups of patches and slides sequentially to capture cross-slide prognostic features. We also present GPAMamba, a model designed to facilitate intra- and inter-slide feature interactions, effectively capturing local micro-environmental characteristics within slide-level graphs while uncovering essential prognostic patterns across an extended patch sequence within the group framework. Furthermore, we develop a dual-head predictor that delivers comprehensive survival risk and probability assessments for each patient. Extensive empirical evaluations demonstrate that our model significantly outperforms state-of-the-art approaches across five datasets from The Cancer Genome Atlas.