[论文解读] Community-Level Modeling of Gyral Folding Patterns for Robust and Anatomically Informed Individualized Brain Mapping

Minheng Chen, Tong Chen+6 more|arXiv (Cornell University)|Feb 1, 2026

Functional Brain Connectivity Studies被引用 0

一句话总结

论文提出一个谱图学习框架，在社区层面建模 gyral folding（3HG folding communities）以实现稳健、解剖学基础的个体化脑映射和跨个体对应的鲁棒性。

ABSTRACT

Cortical folding exhibits substantial inter-individual variability while preserving stable anatomical landmarks that enable fine-scale characterization of cortical organization. Among these, the three-hinge gyrus (3HG) serves as a key folding primitive, showing consistent topology yet meaningful variations in morphology, connectivity, and function. Existing landmark-based methods typically model each 3HG independently, ignoring that 3HGs form higher-order folding communities that capture mesoscale structure. This simplification weakens anatomical representation and makes one-to-one matching sensitive to positional variability and noise. We propose a spectral graph representation learning framework that models community-level folding units rather than isolated landmarks. Each 3HG is encoded using a dual-profile representation combining surface topology and structural connectivity. Subject-specific spectral clustering identifies coherent folding communities, followed by topological refinement to preserve anatomical continuity. For cross-subject correspondence, we introduce Joint Morphological-Geometric Matching, jointly optimizing geometric and morphometric similarity. Across over 1000 Human Connectome Project subjects, the resulting communities show reduced morphometric variance, stronger modular organization, improved hemispheric consistency, and superior alignment compared with atlas-based and landmark-based or embedding-based baselines. These findings demonstrate that community-level modeling provides a robust and anatomically grounded framework for individualized cortical characterization and reliable cross-subject correspondence.

研究动机与目标

• 在保持稳定解剖学标志的同时，解决皮质折叠的显著跨个体变异性。
• 通过对更高阶的折叠社区（3HGs）建模，超越单一标志物匹配。
• 开发一个面向个体的流程，能够在社区层面实现跨个体、解剖学基础的对应。

提出的方法

• 用双重特征描述每个三铰半球回（3HG）：拓扑上下文和结构（trace-map）连通性。
• 在每个受试者半球内使用两层图神经网络对3HG进行聚类，以获得初始社区。
• 应用连接性约束的拓扑细化，确保聚类的空间连续性。
• 通过联合形态–几何匹配（JMGM）在多模态聚类特征（形态、几何和上下文）上用匈牙利分配求解实现跨个体对应。
• 为每个半球分别训练模型以适应个体折叠变异性。

实验结果

研究问题

• RQ1 社区层面的表示是否能改进 gyral folding 模式的跨个体对齐（相对于标志物层面）？
• RQ2 在大型队列中，3HG 折叠社区是否表现出更低的形态测量变异性和更强的模块化结构？
• RQ3 联合形态–几何匹配是否能提供鲁棒的跨个体对应，优于基于图谱或孤立标志的做法？

主要发现

Metric	Left Hemisphere (Range)	Left Hemisphere (Mean ± SD)	Right Hemisphere (Range)	Right Hemisphere (Mean ± SD)	Combined (Mean ± SD)
Node count	[108, 222]	158.97 ± 18.48	[112, 233]	158.96 ± 18.21	158.96 ± 18.35
Edge count	[149, 315]	224.35 ± 18.71	[153, 335]	224.80 ± 18.64	224.57 ± 18.68
Deg-1 nodes	[0, 7]	0.95 ± 0.96	[0, 5]	0.85 ± 0.88	0.90 ± 0.92
Deg-2 nodes	[10, 47]	26.32 ± 5.86	[11, 43]	25.61 ± 5.73	25.96 ± 5.81
Deg-3 nodes	[83, 185]	131.50 ± 16.92	[84, 204]	132.31 ± 16.70	131.91 ± 16.82
Total nodes	167,872	167,860	335,732	335,732	335,732
Total edges	236,916	237,384	474,300	474,300	474,300

• 折叠社区在跨个体的形态测量方差显著降低。
• 社区层面对齐比基于图谱的、现有的标志物或嵌入方法基线在模块化组织和跨个体对齐方面更强。
• 提出的流程提升了半球的一致性和个体特异性折叠表示。
• 跨个体对应在社区层面建立，使个体化脑映射更可靠、粒度更细。
• 该方法在 1,064 名 HCP 受试者中展现出鲁棒性，验证了可扩展性与稳健性。