A Comparative Study of Population-Graph Construction Methods and Graph Neural Networks for Brain Age Regression

Kyriaki Margarita Bintsi*, Tamara T. Mueller, Sophie Starck, Vasileios Baltatzis, Alexander Hammers, Daniel Rueckert

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference paperpeer-review

Abstract

The difference between the chronological and biological brain age of a subject can be an important biomarker for neurodegenerative diseases, thus brain age estimation can be crucial in clinical settings. One way to incorporate multimodal information into this estimation is through population graphs, which combine various types of imaging data and capture the associations among individuals within a population. In medical imaging, population graphs have demonstrated promising results, mostly for classification tasks. In most cases, the graph structure is pre-defined and remains static during training. However, extracting population graphs is a non-trivial task and can significantly impact the performance of Graph Neural Networks (GNNs), which are sensitive to the graph structure. In this work, we highlight the importance of a meaningful graph construction and experiment with different population-graph construction methods and their effect on GNN performance on brain age estimation. We use the homophily metric and graph visualizations to gain valuable quantitative and qualitative insights on the extracted graph structures. For the experimental evaluation, we leverage the UK Biobank dataset, which offers many imaging and non-imaging phenotypes. Our results indicate that architectures highly sensitive to the graph structure, such as Graph Convolutional Network (GCN) and Graph Attention Network (GAT), struggle with low homophily graphs, while other architectures, such as GraphSage and Chebyshev, are more robust across different homophily ratios. We conclude that static graph construction approaches are potentially insufficient for the task of brain age estimation and make recommendations for alternative research directions.

Original languageEnglish
Title of host publicationGraphs in Biomedical Image Analysis, and Overlapped Cell on Tissue Dataset for Histopathology - 5th MICCAI Workshop, GRAIL 2023 and 1st MICCAI Challenge, OCELOT 2023, Held in Conjunction with MICCAI 2023, Proceedings
EditorsSeyed-Ahmad Ahmadi, Sérgio Pereira
PublisherSpringer Science and Business Media Deutschland GmbH
Pages64-73
Number of pages10
ISBN (Print)9783031550874
DOIs
Publication statusPublished - 2024
Event5th Workshop on GRaphs in biomedicAl Image anaLysis Satellite event at MICCAI, GRAIL 2023 and 1st Cell Detection from Cell-Tissue Interaction challenge in MICCAI, OCELOT 2023 Held in Conjunction with International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2023 - Vancouver, Canada
Duration: 8 Oct 202312 Oct 2023

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume14373 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference5th Workshop on GRaphs in biomedicAl Image anaLysis Satellite event at MICCAI, GRAIL 2023 and 1st Cell Detection from Cell-Tissue Interaction challenge in MICCAI, OCELOT 2023 Held in Conjunction with International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2023
Country/TerritoryCanada
CityVancouver
Period8/10/202312/10/2023

Keywords

  • Brain age regression
  • Graph Neural Networks
  • Population graphs

Fingerprint

Dive into the research topics of 'A Comparative Study of Population-Graph Construction Methods and Graph Neural Networks for Brain Age Regression'. Together they form a unique fingerprint.

Cite this