Hierarchical Perfusion Graphs for Tumor Heterogeneity Modeling in Glioma Molecular Subtyping

2026-05-07

AI Summary

This study addresses the limitations of existing structural MRI–based approaches for glioma molecular subtyping, which fail to capture hemodynamic characteristics, and the challenges faced by DSC-MRI radiogenomics due to multicenter variability and voxel-level analysis. To overcome these issues, the authors propose HiPerfGNN, a novel framework that first employs a vector-quantized variational autoencoder (VQ-VAE) to learn discretized perfusion representations from DSC-MRI time-intensity curves, constructing coarse-to-fine functional tumor habitat maps. These are then integrated with structural MRI–guided subregion delineation through a hierarchical graph neural network to enable cross-scale fusion of anatomical and perfusion information. This approach, the first to combine discrete perfusion encoding with structure-guided hierarchical graph modeling, achieves AUCs of 0.96, 0.89, and 0.84 for predicting IDH mutation, 1p/19q codeletion, and WHO grade in an internal cohort (n=475), and maintains robust IDH prediction performance (AUC=0.89) in an external cohort (n=397), while highlighting interpretable regions consistent with pathophysiology.

Abstract

Precise molecular subtyping of gliomas, including isocitrate dehydrogenase (IDH) mutation and 1p/19q codeletion, directly guides surgical and therapeutic decisions, yet currently relies on invasive tissue sampling. Deep learning on structural MRI has emerged as a non-invasive alternative, but anatomy-only approaches cannot capture the hemodynamic signatures that distinguish molecular subtypes. Radiogenomics based on dynamic susceptibility contrast (DSC) MRI holds immense potential for non-invasively characterizing glioma molecular subtypes, yet clinical deployment has been hindered by inter-site variability and the limitations of voxel-wise analysis. We introduce HiPerfGNN, a framework that first learns discrete hemodynamic representations from raw time-intensity curves using a vector-quantized variational autoencoder (VQ-VAE). These quantized perfusion codes define coarse-level graph nodes representing functional tumor habitats, each of which is hierarchically subdivided into fine-level subregions guided by structural MRI. A hierarchical graph neural network then propagates information across scales for molecular prediction. On an internal cohort (n=475), the model achieved AUCs of 0.96 (IDH), 0.89 (1p/19q), and 0.84 (WHO grade), and maintained robust IDH performance (AUC 0.89) on an independent external cohort (n=397) without recalibration. Gradient-based saliency analysis confirms biologically grounded attention patterns aligned with known glioma pathophysiology. Our results demonstrate the added value of integrating perfusion dynamics into radiogenomic pipelines for glioma molecular subtyping. Code is available at https://github.com/janghana/HiPerfGNN.

Problem

• glioma molecular subtyping
• tumor heterogeneity
• DSC-MRI
• radiogenomics
• non-invasive diagnosis

Innovation

• Hierarchical Graph Neural Network
• Vector-Quantized VAE
• Perfusion Radiogenomics
• Tumor Heterogeneity
• DSC-MRI