NeuroGraph

NeuroGraph是由范德堡大学创建的一套用于脑连接组学中图机器学习基准的数据集。该数据集包含35个子数据集，涵盖了静态和动态脑连接数据，用于预测多种行为和认知特征。数据集的创建过程涉及从功能磁共振成像（fMRI）数据中提取区域兴趣（ROI）的时间序列，进行预处理和构建静态及动态脑图。NeuroGraph的应用领域包括性别、年龄、任务分类以及流体智力和工作记忆得分的预测，旨在解决脑连接分析中的复杂性和参数空间大的问题。

NeuroGraph is a benchmark dataset for graph machine learning in brain connectomics, developed by Vanderbilt University. This dataset consists of 35 sub-datasets covering both static and dynamic brain connectivity data, and is utilized for predicting various behavioral and cognitive traits. The development process of NeuroGraph involves extracting time series of regions of interest (ROI) from functional magnetic resonance imaging (fMRI) data, performing preprocessing, and constructing both static and dynamic brain graphs. The application scenarios of NeuroGraph include gender, age and task classification, as well as the prediction of fluid intelligence and working memory scores. It aims to address the challenges of complexity and large parameter space in brain connectivity analysis.

搜集汇总

背景与挑战

背景概述

NeuroGraph是一个基于HCP Young Adult dataset的图机器学习神经影像数据集集合，涵盖性别、年龄分类、心理状态解码以及流体智力和工作记忆预测等多类认知和行为任务。该数据集提供静态和动态图表示，适用于fMRI预处理和图机器学习应用，旨在支持脑连接组学的研究。