F-TRACT: a probabilistic atlas of anatomo-functional connectivity of the human brain (F-TRACT_P_10_v2210)

These data come from a cohort of adult patients suffering from pharmaco-resistant focal epilepsies who in the course of preparation to a brain surgery were implanted with intracerebral stereoencephalographic (SEEG) electrodes. The protocol involved brain stimulation at a single site and a simultaneous recording with all remaining contacts on a number of electrodes. Beyond medical aim, performing a number of such stimulations in a single patient allows to infer partial information about their brain anatomo-functional connectivity. What is more, aggregating data from a large cohort of such patients, thus from implantations at different brain regions, allows to derive probabilistic connectivity of the whole human brain. This was achieved by the Functional Tractography project ((F-TRACT)[https://f-tract.eu]) and the results are shared in this dataset. The aggregated data are projected onto a number of parcellations. They describe probabilistic connectivity and connectivity features (such as signal propagation delay) in a context of two parcels: the stimulated and the recorded one. A map for one feature in one parcellation is expressed as a square matrix that can be asymmetric. These matrices were derived from signal analysis of responses to stimulation on an individual level and then aggregation of results on the group level. They can be used to inform about connectivity strength and latency of signal propagation between different areas of the human brain. The naming convention for the atlas versions is the following: "F-TRACT\_P\_XX\_vYYMM" "F-TRACT" and "P" are the common across versions. "F-TRACT" stands for "functional tractography" and "P" stands for "parcellation-based". "XX" varies across alternative versions of a major atlas release and refers to different quality assessments of the same source data as binary code. "vYYMM" varies only across major atlas releases where "YYMM" refers to the year and month of the release. Note that major atlas releases can differ in the source data and/or in the selection of the applied parcellations.

本数据集的样本来自一组罹患药物难治性局灶性癫痫（pharmaco-resistant focal epilepsies）的成年患者：这些患者在接受脑外科手术术前准备阶段，被植入了颅内立体脑电图（SEEG）电极。本研究方案包含单点位脑刺激，以及对多根电极上其余所有触点进行同步记录。除临床诊疗目标外，对单名患者实施多次此类刺激，可推断其大脑解剖-功能连接的部分信息。进一步而言，整合来自该大队列患者（即植入位置覆盖不同脑区的患者）的数据，可推导全人类大脑的概率性连接模式。该成果由功能束追踪（Functional Tractography）项目（简称F-TRACT，网址：https://f-tract.eu）完成，相关结果已收录于本数据集。整合后的数据被投影至多种脑区划分方案（parcellation）。这些数据描述了两类脑区（刺激位点与记录位点）之间的概率性连接模式及连接特征（如信号传播延迟）。某一脑区划分方案下的单一项特征图谱，可表示为非对称方阵。此类方阵源自个体水平上对刺激响应的信号分析，随后经群体水平的结果整合得到。该类数据可用于阐释人类大脑不同脑区间的连接强度与信号传播潜伏期。该脑图谱版本的命名规则如下：格式为"F-TRACT_P_XX_vYYMM"。其中"F-TRACT"与"P"为所有版本通用前缀："F-TRACT"代表「功能束追踪」，"P"代表「基于脑区划分」。"XX"用于区分同一正式图谱发布版本的不同分支，以二进制编码形式表示对同一源数据的不同质量评估结果。"vYYMM"仅用于区分不同正式图谱发布版本，其中"YYMM"代表该版本的发布年月。需注意，不同正式图谱发布版本可能在源数据或所采用的脑区划分方案选择上存在差异。