融合拓扑与图结构，用对比学习分析神经元形态
GraPHFormer: A Multimodal Graph Persistent Homology Transformer for the Analysis of Neuroscience Morphologies
- 用三通道持久性图像和树LSTM联合建模拓扑与几何特征
- 在六个数据集上五项超越当前最佳，显著优于单一视角方法
- 适合神经形态分析、发育与退行性疾病研究者使用
神经元形态编码了电路功能、发育和疾病的关键信息，但现有方法通常孤立地分析拓扑或图结构。我们提出GraPHFormer，一种多模态架构，通过类似CLIP的对比学习统一这两种互补视角。视觉分支利用DINOv2-ViT-S处理新型三通道持久性图像，编码无权、持久性加权和半径加权的拓扑密度；并行的TreeLSTM编码器捕获骨架图的几何与径向属性。两者映射到共享嵌入空间，通过对称InfoNCE损失训练，并引入保持拓扑语义的持久性空间变换。在六个基准（BIL-6, ACT-4, JML-4, N7, M1-Cell, M1-REG）上评估，涵盖自监督与监督设置，GraPHFormer在五个基准上达到当前最优性能，显著优于仅拓扑、仅图结构及形态度量基线。我们验证其实际价值：可区分不同皮层区域和物种的胶质细胞形态，并检测发育与退行性过程的特征。代码已开源。
原文摘要 · Abstract (English)
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
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。