遇见数据集

Ginkgo Herlin's Deep White Matter Atlas of the Human Brain

收藏
Zenodo2025-02-21 更新2026-05-26 收录
官方服务:

资源简介:

The Deep White Matter Atlas of the Human Brain was built on the entire HCP-YA cohort, made of 1065 young healthy subjects. The construction of this atlas is based on the analysis of the anatomical and diffusion MRI acquisitions of each subject, using the tractography and fiber clustering tools available in the Ginkgo toolbox (CEA, NeuroSpin, BAOBAB, GAIA, Ginkgo Team, https://framagit.org/cpoupon/gkg). Detailed information on the construction of the atlas can be found in the related article. The atlas can be visualized with the BrainVISA/Anatomist viewer available at https://brainvisa.info/web/download.html. The atlas is registered in the MNI space. This atlas is composed of 77 white matter tracts, including: - 30 association tracts (15 per hemisphere): arcuate fasciculus, corticospinal tract, dorsal cingulum, ventral cingulum, long dorsoventral cingulum, external capsule, extreme capsule, frontal aslant, inferior fronto-occipital fasciculus, inferior longitudinal fasciculus, middle longitudinal fasciculus, superior longitudinal fasciculus 1, superior longitudinal fasciculus 2, superior longitudinal fasciculus 3, and uncinate fasciculus. - 38 projection tracts (19 per hemisphere): fornix, centro-caudate tract, cingulo-caudate tract, fronto-caudate tract, parieto-caudate tract, centro-lenticular tract, fronto-lenticular tract, occipito-lenticular tract, parieto-lenticular tract, temporo-lenticular tract, spino-cerebellar tract, middle cortico-cerebellar tract, superior cortico-cerebellar tract, optic radiation, thalamo-central tract, thalamo-frontal tract, thalamo-occipital tract, thalamo-parietal tract, and thalamo-temporal tract. - 8 inter-hemispheric tracts : anterior comissure, and 7 subdivisions of the corpus callosum according to Wittelsion segmentation (rostrum, genu, rostral body, anterior midbody, posterior midbody, isthmus, splenium) - 1 intra-cerebellar tract : the parallel fibers. The atlas is provided in the Anatomist *.bundles/*.bundlesdata format, for which the *.bundles file contains meta information, including: - The labels of the different white matter bundles ('labels' entry), - The number of streamlines populating each white matter bundle ('curve3d_counts' entry), in the same order as the 'labels' key, - The total number of white matter bundles ('item_count' entry), - The total number of streamlines ('curves_count' entry). Two different versions of the atlas are provided : - A non-symmetrized version of the atlas in which the left and right tracts retain the anatomical differences found within the HCP-YA cohort. This version is better suited for visual representation of deep white matter tracts. This atlas is referenced as "Ginkgo_Herlin_HumanDeepWhiteMatterAtlas_Unsymmetrized". - A symmetrized version of the atlas in which the left and right tracts have been symmetrized, i.e. the left and right tracts are exactly the same with respect to an axis of symmetry along the sagital plane, and the interhemispheric tracts have been fused with their symmetric versions. This version is more suitable for tractography analysis: the atlas can be applied to a whole-brain tractogram of any subject (registered in MNI space) to automatically label the deep white matter tracts, for example using a minimum distance threshold algorithm (with the optimized distance threshold for each tract specified in the maximum-distance-to-bundle-map.py file). The symmetric version of the atlas ensures that any left-right difference found after labeling is due to a difference within the subject's tractogram and not to a difference in the atlas itself. This atlas is referenced as "Ginkgo_Herlin_HumanDeepWhiteMatterAtlas_Symmetrized". Each atlas is provided in a high-resolution version with a fiber resampling step of 1mm, and a low-resolution version with a fiber resampling step of 3mm. This resampling step is specified at the end of the atlas filename as "_1mm" and "_3mm" respectively.

提供机构:
Zenodo
创建时间:
2025-02-21
二维码
社区交流群
二维码
科研交流群
商业服务