DiFuMo atlas (256 dimensions)

DiFuMo (Dictionary of Functional Modes) 256-dimensional atlas presents a parcellation of the brain into 256 elementary structures using sparse dictionary learning. The atlas contains two main components: (1) an image of the 256 spatial components, sampled in MNI space, and (2) anatomical names for these 256 regions. It is meant to serve as an atlas to extract functional signals, eg to create image-derived phenotypes. These modes are optimized to represent well raw BOLD time series, over a with range of experimental conditions.

DiFuMo（功能模式字典，Dictionary of Functional Modes）256维脑图谱是一种通过稀疏字典学习算法将大脑划分为256个基本结构的脑分区图谱。该图谱包含两大核心组成部分：(1) 在MNI空间中采样得到的256个空间成分图像；(2) 这256个脑区的解剖学命名。其设计初衷是作为可提取功能信号的标准化图谱，例如用于构建影像衍生表型（image-derived phenotypes）。这些功能模式经过优化，可在多种实验条件下精准表征原始血氧水平依赖（Blood Oxygen Level Dependent, BOLD）时间序列。