Skip to main content
QUICK REVIEW

[论文解读] PTGB: Pre-Train Graph Neural Networks for Brain Network Analysis

Yi Yang, Hejie Cui|arXiv (Cornell University)|May 20, 2023
Functional Brain Connectivity Studies被引用 4
一句话总结

PTGB 是一种用于脑网络分析的自监督图神经网络预训练框架,通过两级对比学习和数据驱动的脑区图谱映射流程,解决了数据稀缺和不同脑部分割图谱不一致的问题。该方法在多个脑部疾病预测任务中达到最先进性能,在关键指标上相比基线模型相对提升4.27%。

ABSTRACT

The human brain is the central hub of the neurobiological system, controlling behavior and cognition in complex ways. Recent advances in neuroscience and neuroimaging analysis have shown a growing interest in the interactions between brain regions of interest (ROIs) and their impact on neural development and disorder diagnosis. As a powerful deep model for analyzing graph-structured data, Graph Neural Networks (GNNs) have been applied for brain network analysis. However, training deep models requires large amounts of labeled data, which is often scarce in brain network datasets due to the complexities of data acquisition and sharing restrictions. To make the most out of available training data, we propose PTGB, a GNN pre-training framework that captures intrinsic brain network structures, regardless of clinical outcomes, and is easily adaptable to various downstream tasks. PTGB comprises two key components: (1) an unsupervised pre-training technique designed specifically for brain networks, which enables learning from large-scale datasets without task-specific labels; (2) a data-driven parcellation atlas mapping pipeline that facilitates knowledge transfer across datasets with different ROI systems. Extensive evaluations using various GNN models have demonstrated the robust and superior performance of PTGB compared to baseline methods.

研究动机与目标

  • 为解决脑网络分析中标签数据有限的问题,该问题限制了深度学习的性能表现。
  • 克服不同神经影像数据集中感兴趣脑区(ROI)分割系统不一致的问题,该问题阻碍了知识迁移。
  • 开发一种预训练框架,从无任务标签的脑网络中学习结构归纳偏置。
  • 通过学习不同脑部分割图谱之间的共享、可解释的映射关系,实现有效的跨数据集适应。
  • 通过联合优化图级和节点级表征,提升下游疾病预测性能。

提出的方法

  • 提出一种两级节点对比采样策略,同时优化局部邻域和全局图级表征。
  • 引入基于多图脑网络间互信息最大化构建的多数据集自监督预训练目标。
  • 采用数据驱动的图谱映射流程,使用三种专用正则化项:稀疏性、局部性和社区正则化项,实现不同分割模板间脑区的对齐。
  • 利用方差排序方法识别并优先选择最具信息量的虚拟脑区,以实现跨图谱对齐。
  • 应用受MAML启发的元学习原则,实现预训练模型对下游任务的快速适应。
  • 集成图级对比学习模块,将全局图结构信息总结为局部节点表征。
Figure 1: Overview of the proposed framework PTGB . The initial features of the source datasets are projected to a fixed dimension through atlas transformation followed by variance-based feature alignment, which facilitates self-supervised GNN pre-training on multiple datasets via the novel two-leve
Figure 1: Overview of the proposed framework PTGB . The initial features of the source datasets are projected to a fixed dimension through atlas transformation followed by variance-based feature alignment, which facilitates self-supervised GNN pre-training on multiple datasets via the novel two-leve

实验结果

研究问题

  • RQ1在大规模无标签脑网络上进行自监督预训练,是否能提升下游疾病预测性能?
  • RQ2两级对比学习在捕捉脑网络中局部与全局结构模式方面的有效性如何?
  • RQ3在不依赖人工标注的前提下,数据驱动的图谱映射流程在多大程度上能实现不同分割系统间脑区的对齐?
  • RQ4PTGB框架中各组件——特别是正则化项与对比学习目标——对整体性能的贡献如何?
  • RQ5PTGB在具有不同脑区定义和临床结局的多样化数据集上,泛化能力如何?

主要发现

  • 完整PTGB框架在准确率和AUC两个指标上均较最强基线模型实现4.27%的相对提升,证明了在多张图中联合学习共享子结构知识的价值。
  • 移除稀疏性正则化项导致性能下降最大,证实其在提升模型鲁棒性和防止过拟合方面具有关键作用。
  • 引入局部性和社区正则化项显著提升性能,验证了对空间结构和模块化脑网络结构认知的重要性。
  • 两级对比学习(节点级与图级)相比单级方法收敛更快,且在下游适应任务中表现更优。
  • 基于方差的ROI对齐方法成功识别出PPMI、BP和HIV数据集中重叠的脑区,可视化结果确认了区域间的对应关系。
  • 该框架在三个不同数据集(PPMI、BP和HIV)上均保持优异性能,证明其在多样化临床与影像场景下的鲁棒性与泛化能力。
Figure 2: Visual demonstration of the sample types where $X_{i,p}$ is the anchor and $\mathbf{S_{1}}$ / $\mathbf{S_{4}}$ are sampled as 1-hop neighbors.
Figure 2: Visual demonstration of the sample types where $X_{i,p}$ is the anchor and $\mathbf{S_{1}}$ / $\mathbf{S_{4}}$ are sampled as 1-hop neighbors.

更好的研究,从现在开始

从阅读论文到最终审阅,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。