PuSH - Publication Server of Helmholtz Zentrum München

Demirbilek, O.* ; Peng, T. ; Bessadok, A.

Graph residual noise learner network for brain connectivity graph prediction.

In: (Graphs in Biomedical Image Analysis). Berlin [u.a.]: Springer, 2025. 23-32 (Lect. Notes Comput. Sc. ; 15182)
A morphological brain graph depicting a connectional fingerprint is of paramount importance for charting brain dysconnectivity patterns. Such data often has missing observations due to various reasons such as time-consuming and incomplete neuroimage processing pipelines. Thus, predicting a target brain graph from a source graph is crucial for better diagnosing neurological disorders with minimal data acquisition resources. Many brain graph generative models were proposed for promising results, yet they are mostly based on generative adversarial networks (GAN), which could suffer from mode collapse and require large training datasets. Recent developments in diffusion models address these problems by offering essential properties such as a stable training objective and easy scalability. However, applying a diffusion process to graph edges fails to maintain the topological symmetry of the brain connectivity matrices. To meet these challenges, we propose the Graph Residual Noise Learner Network (Grenol-Net), the first graph diffusion model for predicting a target graph from a source graph. Its two core contributions lie in (i) introducing a graph diffusion model that learns nodelevel noise for accurate denoising (ii) introducing a node-based diffusion function to better maintain the topological structure of brain graphs. Our Grenol-Net is composed of graph convolutional blocks, which first learn the source embeddings and second, a set of fully connected layers assisted with a positional encoding block that predicts the nodes with noise level t- 1 in the target domain. We further design a batch normalization block that learns the target distribution at diffusion timestep t and operates an element-wise subtraction from the predicted nodes with noise level t - 1. Our Grenol-Net outperformed existing methods on the morphological brain graph extracted from cortical measurements of the left and right hemispheres separately and on three distinct datasets from multiple cohorts.
Additional Metrics?
[➜Log in]
Edit extra informations Login
Publication type Article: Conference contribution
Keywords diffusion models; graph neural networks; brain graph prediction; morphological brain network
ISSN (print) / ISBN 0302-9743
e-ISSN 1611-3349
Conference Title Graphs in Biomedical Image Analysis
Journal Lecture Notes in Computer Science
Quellenangaben Volume: 15182, Issue: , Pages: 23-32 Article Number: , Supplement: ,
Publisher Springer
Publishing Place Berlin [u.a.]
Institute(s) Helmholtz Artifical Intelligence Cooperation Unit (HAICU)
Grants Humboldt Postdoctoral Research Fellowship