Gast ::
| Hauptseite > Publikationsdatenbank > Deep texture features characterizing fiber architecture in the vervet monkey occipital lobe (v1) |
| Hauptseite > Publikationsdatenbank > Deep texture features characterizing fiber architecture in the vervet monkey occipital lobe (v1) |
| Dataset | FZJ-2024-07234 |
; ; ; ; ; ; ; ; ;
2024
EBRAINS
This record in other databases:
Please use a persistent id in citations: doi:10.25493/78A4-KTU
Abstract: This dataset contains features for characterizing fiber architecture in three-dimensional polarized light imaging (3D-PLI) for the right occipital lobe from a single vervet monkey brain, comprising 234 brain sections. The data include volumetric PCA projections of deep texture features, clusterings, as well as measures of cortical morphology, such as curvature, cortical depth, white matter depth, and section obliqueness. PCA projections form 20-dimensional vectors, each representing a square image patch of 169 μm, sampled at a particular section and location in the brain. Morphological features were obtained through a cortex segmentation of the volume to explore correlations between extracted features and morphology. Deep texture features were extracted using a self-supervised 3D-Context Contrastive Learning (CL-3D) model, trained on high-resolution microscopic 3D-PLI images. These features reveal patterns of fiber architecture, including distinctions between gray and white matter, myeloarchitectonic layer structures, fiber bundles, fiber crossings, and fiber fannings. Of the 234 sections, 117 were used to train the feature extraction model, and the remaining 117 are provided for analysis. All volumes are spatially aligned with the Average MRI Vervet Atlas of the UCLA Brain Mapping Center using affine matrices. The alignment allows each texture feature to be assigned to a 3D coordinate in this space, referring to the centroid of the patch, for comparison with other data.
Keyword(s): Neuroscience
|
The record appears in these collections: |