Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Graph Residual Noise Learner Network for Brain Connectivity Graph Prediction

Oytun Demirbilek, Tingying Peng, Alaa Bessadok

TL;DR

The Graph Residual Noise Learner Network (Grenol-Net) is proposed, the first graph diffusion model for predicting a target graph from a source graph and fails to maintain the topological symmetry of the brain connectivity matrices.

Abstract

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.

Graph Residual Noise Learner Network for Brain Connectivity Graph Prediction

TL;DR

The Graph Residual Noise Learner Network (Grenol-Net) is proposed, the first graph diffusion model for predicting a target graph from a source graph and fails to maintain the topological symmetry of the brain connectivity matrices.

Abstract

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.
Paper Structure (9 sections, 8 equations, 3 figures, 1 table)

This paper contains 9 sections, 8 equations, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Methodology
  3. Experiments and Results
  4. Datasets
  5. Hyperparameter tuning
  6. Benchmarks
  7. Evaluation
  8. Conclusion
  9. Acknowledgement

Figures (3)

  • Figure 1: Overview of our Grenol-Net architecture for predicting a target brain graph from source graph by learning conditional denoising process.
  • Figure 2: Evaluation in terms of Frobenius distance and mean squared error (MSE) of the respective model MGCN-GAN zhang2022predicting, DGN gurbuzrekik2020, HADA bessadok2019hierarchical, and Grenol-Net separately on different hemispheres.
  • Figure 3: Comparison in terms of Frobenius distance of DGN gurbuzrekik2020 and Grenol-Net trained on Openneuro dataset and tested on cross-cohort datasets.