Texture-based approaches for identifying neuro-anatomical structures and electrode tracks |
| |
Authors: | Xiang Yongqing Büttner-Ennever Jean Cohen Bernard Raphan Theodore |
| |
Affiliation: | Department of Computer and Information Science, Brooklyn College of the City University of New York, 2900 Bedford Avenue, Brooklyn, NY 11210, USA. |
| |
Abstract: | An automated approach to identifying electrode tracks and neuro-anatomical structures (nuclei) was developed using texture attributes of their neuro-anatomical stains. The properties that make up the texture features of the nuclei include size, shape and distribution of elemental structures. The electrode tracks are characterized by elongated darkened formations due to gliosis. Based on a Gabor wavelet transform, a texture feature vector was constructed, consisting of localized texture energies along different orientations at different scales. Stained images of brainstem sections in the vestibular nuclei were segmented using partitional clustering in feature space. A metric that computes the location of the tracks relative to the nuclei centers was then implemented. This methodology should be useful for quantifying and automating the procedure by which tracks are localized in anatomical structures. |
| |
Keywords: | |
本文献已被 PubMed 等数据库收录! |
|