Spatiotemporal Graph Neural Network Modelling Perfusion MRI

TL;DR

This paper introduces PerfGAT, a novel spatiotemporal graph neural network model designed to analyze 4D perfusion MRI data for predicting IDH mutation status in glioma patients, addressing challenges in modeling temporal correlations and data imbalance.

Contribution

The study presents the first GNN-based model for 4D pMRI, incorporating edge attention, negative graphs, dual-attention fusion, and class-balanced augmentation to improve tumor characterization.

Findings

Outperforms existing methods in IDH mutation prediction

Effectively models spatiotemporal features of perfusion MRI

Mitigates label imbalance in clinical datasets

Abstract

Perfusion MRI (pMRI) offers valuable insights into tumor vascularity and promises to predict tumor genotypes, thus benefiting prognosis for glioma patients, yet effective models tailored to 4D pMRI are still lacking. This study presents the first attempt to model 4D pMRI using a GNN-based spatiotemporal model PerfGAT, integrating spatial information and temporal kinetics to predict Isocitrate DeHydrogenase (IDH) mutation status in glioma patients. Specifically, we propose a graph structure learning approach based on edge attention and negative graphs to optimize temporal correlations modeling. Moreover, we design a dual-attention feature fusion module to integrate spatiotemporal features while addressing tumor-related brain regions. Further, we develop a class-balanced augmentation methods tailored to spatiotemporal data, which could mitigate the common label imbalance issue in clinical datasets. Our experimental results demonstrate that the proposed method outperforms other state-of-the-art approaches, promising to model pMRI effectively for patient characterization.