Semi-automatic brain region segmentation method for three-dimensional cell construction image

Abstract

The invention provides a semi-automatic brain region segmentation method for a three-dimensional cell construction image. The semi-automatic brain region segmentation method comprises the steps of S1,knowledge introduction; s2, knowledge digitization; s3, knowledge packaging; s4, brain area automatic identification; and S5, optimizing of the boundary of the brain area. According to the method, higher-dimensional and more abstract features can be better extracted through a deep learning method, so that the overall boundary of a brain region formed by discrete cell bodies can be identified; further, through interactive segmentation, according to the invention, priori knowledge of neural anatomy is successfully introduced into a prediction network constructed by a deep learning technology; neuroanatomy priori knowledge mastered by experts is digitized; by means of the method, the experience which is only expected to be difficult to speak in the brain of an expert can be changed into a tool to be repeatedly used, the threshold of automatic recognition of the brain area is greatly lowered, so that a common neuroscience researcher gets rid of high dependence on knowledge in the narrow field of neuroanatomy, and the recognition efficiency of the brain area is greatly improved.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a semi-automatic brain region segmentation method for three-dimensional cell construction images at a resolution level of hundreds of microns to microns. Background technique [0002] Brain imaging is one of the necessary technical means to carry out neuroscience research. With the advancement of microscopic optical imaging technology, people have obtained a large amount of three-dimensional brain image data at the micron resolution level, laying the foundation for research on a finer spatial scale. Among the many types of brain images, cytoarchitecture images are considered the gold standard for identifying different brain regions in the brain. On this type of image, the cell body shape and spatial aggregation pattern of cells in different brain regions are different from those visible to the human eye. Therefore, experienced neuroanatomy experts usually find and manually draw t...

AI Technical Summary

Problems solved by technology

For the extraction of such area boundaries, traditional image segmentation algorithms that rely on continuous and regular grayscale information are helpless

Images