Survival Modeling from Whole Slide Images via Patch-Level Graph Clustering and Mixture Density Experts

TL;DR

This paper introduces a modular framework that leverages patch-level graph clustering and mixture density models to predict cancer survival from whole slide images, capturing heterogeneity at multiple scales.

Contribution

It presents a novel multi-stage approach combining patch filtering, graph clustering, hierarchical feature aggregation, and mixture density modeling for survival prediction from WSIs.

Findings

Achieved state-of-the-art concordance indices on TCGA cohorts.

Effectively models tumor heterogeneity and spatial relationships.

Outperforms existing methods in survival prediction accuracy.

Abstract

We propose a modular framework for predicting cancer specific survival directly from whole slide pathology images (WSIs). The framework consists of four key stages designed to capture prognostic and morphological heterogeneity. First, a Quantile Based Patch Filtering module selects prognostically informative tissue regions through quantile thresholding. Second, Graph Regularized Patch Clustering models phenotype level variations using a k nearest neighbor graph that enforces spatial and morphological coherence. Third, Hierarchical Feature Aggregation learns both intra and inter cluster dependencies to represent multiscale tumor organization. Finally, an Expert Guided Mixture Density Model estimates complex survival distributions via Gaussian mixtures, enabling fine grained risk prediction. Evaluated on TCGA LUAD, TCGA KIRC, and TCGA BRCA cohorts, our model achieves concordance indices of 0.653 ,0.719 ,and 0.733 respectively, surpassing existing state of the art approaches in survival prediction from WSIs.