GraPHFormer: A Multimodal Graph Persistent Homology Transformer for the Analysis of Neuroscience Morphologies

2026-03-21

AI Summary

Existing approaches struggle to jointly model the topological and geometric structures of neuronal morphology, hindering a deeper understanding of neural circuit function, development, and disease mechanisms. This work proposes GraPHFormer, the first framework to integrate persistence homology images and skeleton graphs through CLIP-style contrastive learning. It leverages DINOv2-ViT-S to extract three-channel persistence homology features, employs TreeLSTM to encode graph structure, and introduces a topology-aware augmentation strategy that preserves topological semantics, enabling multimodal alignment in a shared embedding space. Evaluated on six neuroanatomical benchmarks, GraPHFormer achieves state-of-the-art performance on five, significantly outperforming methods based solely on topology, graph structure, or traditional morphometrics. The model successfully uncovers cross-regional and cross-species differences in glial cells and effectively identifies key features associated with developmental and degenerative changes.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Computer Vision: Other Foundations of Computer Vision

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Neuronal morphology encodes critical information about circuit function, development, and disease, yet current methods analyze topology or graph structure in isolation. We introduce GraPHFormer, a multimodal architecture that unifies these complementary views through CLIP-style contrastive learning. Our vision branch processes a novel three-channel persistence image encoding unweighted, persistence-weighted, and radius-weighted topological densities via DINOv2-ViT-S. In parallel, a TreeLSTM encoder captures geometric and radial attributes from skeleton graphs. Both project to a shared embedding space trained with symmetric InfoNCE loss, augmented by persistence-space transformations that preserve topological semantics. Evaluated on six benchmarks (BIL-6, ACT-4, JML-4, N7, M1-Cell, M1-REG) spanning self-supervised and supervised settings, GraPHFormer achieves state-of-the-art performance on five benchmarks, significantly outperforming topology-only, graph-only, and morphometrics baselines. We demonstrate practical utility by discriminating glial morphologies across cortical regions and species, and detecting signatures of developmental and degenerative processes. Code: https://github.com/Uzshah/GraPHFormer

Problem

• neuronal morphology
• topological analysis
• graph structure
• multimodal integration
• neuroscience

Innovation

• Graph Persistent Homology
• Multimodal Contrastive Learning
• Neuronal Morphology Analysis
• TreeLSTM
• Topological Data Analysis