Thalamus nucleus partitioning method based on global connection characteristics and geodesic distance

Abstract

The invention discloses a thalamus nucleus partitioning method based on global connection characteristics and geodesic distance so as to realize brain thalamus area nucleus partitioning. The method comprises the following steps that: firstly, utilizing the time sequence information of a brain resting state magnetic resonance image, combining with the priori atlas of different tissues of the brainto construct the global connection characteristics of a thalamus voxel level, and lowering characteristic dimensions to reduce noise; then, combining with the three-dimensional space structure information of the brain, constructing a thalamus similarity matrix based on the geodesic distance, and more effectively utilizing the information of the characteristics; and finally, putting forward a normalization spectrum segmentation clustering method to combine with thalamus voxel similarity information to cluster the voxel. Compared with a traditional method, the method which is put forward by theinvention can more effectively lower the noise influence of the resting state magnetic resonance image, high-quality characteristics are extracted, and an accurate and effective thalamus nucleus partition is generated.

Description

technical field [0001] The invention relates to a method for partitioning thalamic nuclei, in particular to a method for partitioning thalamic nuclei based on global connection features and geodesic distances, and belongs to the technical field of digital images. Background technique [0002] The thalamus is the largest oval gray matter nucleus in the human diencephalon. In animals with underdeveloped cerebral cortex, the thalamus is the highest-level sensory center; in animals with developed cerebral cortex, it is the most important sensory transmission relay station, and all kinds of sensations (except smell) from the whole body can only be transmitted through the thalamus. to the cortical area. Dysregulation of the thalamus has been implicated in physiological and neurological disorders. At the cytoarchitectural level, the thalamus can be divided into several subregions, each with different connections to the cortex, so that different segmentations of the thalamus can d...

AI Technical Summary

Problems solved by technology

In the traditional method, the use of rs-fMRI to partition the thalamus mainly uses the prior seed point method based on functional connectivity. This method requires prior partition knowledge of the cortex. The currently known cortical templates include Automated Anatomical Labeling (AAL ) template, Harvard-Oxford (HO) template, etc., Fan et al. used the time series correlation between thalamic voxels as a feature to cluster the thalamus, but only used the local information of thalamic voxels without considering the thalamus and cortical brain Connectivity between districts

Images