A method for constructing a probability atlas of human gray matter nuclei based on quantitative magnetic susceptibility imaging

Abstract

The invention discloses a human brain gray matter nucleus probability map construction method based on quantitative magnetic susceptibility imaging. The method comprises steps that magnetic resonance imaging equipment is utilized to acquire data; a detected quantitative magnetic susceptibility image is registered to a standard space, namely an MNI space; a brain deep gray matter nucleus region is sketched manually on the magnetic susceptibility image of the standard space after registering, and different probability maps are made according to different overlapping proportions; for the magnetic susceptibility image for evaluation, similarity and coverage rate value measurement on an automatic segmentation result of different probability maps and a golden standard acquired through manual sketching is carried out, a map of an corresponding overlapping proportion is employed to construct a final probability map when similarity reaches a peak value. The method is advantaged in that the constructed probability map realizes automatic segmentation of the brain deep gray matter nucleus to avoid artificial errors caused by manual sketch, the nucleus segmentation result has relatively high accuracy, the method is better than AAL and JH maps in the prior art, moreover, the time is better saved compared with a manual sketch method, and image analysis work efficiency is improved.

Description

technical field [0001] The invention belongs to the technical field of magnetic resonance imaging, in particular to a method for constructing a probabilistic map of human brain deep gray matter nuclei based on quantitative magnetic susceptibility imaging. Background technique [0002] The three-dimensional brain atlas can provide a useful anatomical reference for the analysis of brain imaging data. Through the automatic registration between the subject and the brain template, the brain atlas can effectively segment the subject's brain into corresponding regions of interest. The structure and function of the human brain are complex and diverse, and each part has its specificity. Spatial normalization can reduce the differences in brain anatomy between individuals, so it is an important step in the study of human brain atlas. At present, brain atlases widely used in magnetic resonance research are derived from T1-weighted images, such as Talairach and Tournoux atlases, automa...

AI Technical Summary

Problems solved by technology

Recently, a research team from Johns Hopkins University produced a brain atlas (JH atlas) based on quantitative magnetic susceptibility maps, but the atlas is based on a single subject's brain and cannot fully reflect the anatomical diversity of a large number of normal brains

Images