研究人员开发了一种名为“大脑图谱上的双曲学习”（HLBG）的新型框架，用于分析功能性大脑网络和诊断疾病。该方法利用深度双曲学习来模拟大脑网络固有的层级结构，从个体区域（ROIs）到全脑整合。HLBG将表示投影到洛伦兹双曲空间，并结合了图感知Mamba（GaMamba）模型，以捕捉长距离依赖性并保留拓扑信息。在ABIDE-I和REST-MDD数据集上的实验表明，HLBG优于现有的最先进方法，并识别出与疾病相关的生物标志物。 AI
影响 通过改进对复杂大脑网络数据的分析，这项研究可能带来更准确、更有效的神经和精神疾病诊断工具。
排序理由 该集群包含一篇详细介绍大脑网络分析新方法的学术论文。
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
arXiv:2607.07077v1 Announce Type: cross Abstract: Functional brain networks exhibit a hierarchical organization across ROI, community, and whole-brain levels, supporting local processing, inter-community coordination, and global integration. Recent studies have demonstrated that …
Functional brain networks exhibit a hierarchical organization across ROI, community, and whole-brain levels, supporting local processing, inter-community coordination, and global integration. Recent studies have demonstrated that brain community-aware modeling is beneficial for b…
Functional brain networks exhibit a hierarchical organization across ROI, community, and whole-brain levels, supporting local processing, inter-community coordination, and global integration. Recent studies have demonstrated that brain community-aware modeling is beneficial for b…