Unified Generative Latent Representation for Functional Brain Graphs

2025-11-06

AI Summary

This study addresses the challenge of learning a unified low-dimensional latent representation for functional brain connectomes that jointly captures topological and spectral structural covariations—without relying on isolated graph-theoretic or spectral metrics. We propose a geometry-aware unsupervised generative framework: a Graph Transformer encoder integrates spectral-geometric priors as inductive bias, coupled with a latent diffusion model to enable high-fidelity reconstruction. The method embeds functional connectivity dynamics onto a low-dimensional manifold that reflects continuous geometric variations of brain networks. It effectively discriminates working memory states, decodes visual stimuli, and synthesizes structurally coherent, biologically interpretable functional connectomes. Incorporating neural dynamical information further enhances both generative fidelity and downstream task performance—demonstrating improved classification accuracy and stimulus decoding reliability.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Functional brain graphs are often characterized with separate graph-theoretic or spectral descriptors, overlooking how these properties covary and partially overlap across brains and conditions. We anticipate that dense, weighted functional connectivity graphs occupy a low-dimensional latent geometry along which both topological and spectral structures display graded variations. Here, we estimated this unified graph representation and enabled generation of dense functional brain graphs through a graph transformer autoencoder with latent diffusion, with spectral geometry providing an inductive bias to guide learning. This geometry-aware latent representation, although unsupervised, meaningfully separated working-memory states and decoded visual stimuli, with performance further enhanced by incorporating neural dynamics. From the diffusion modeled distribution, we were able to sample biologically plausible and structurally grounded synthetic dense graphs.

Problem

• Unifying disjoint graph descriptors into a shared latent space
• Generating biologically plausible functional brain connectivity graphs
• Capturing graded variations in topological and spectral structures

Innovation

• Graph transformer autoencoder with latent diffusion
• Spectral geometry guides unsupervised learning
• Generates biologically plausible synthetic brain graphs