Choroid vessel segmentation method and system based on three-dimensional coherent tomographic image

Abstract

The invention discloses a choroid vessel segmentation method and system based on a three-dimensional coherent tomographic image. The method includes the steps of collecting the three-dimensional frequency domain coherent optical tomographic image; calculating a fundus radian model according to the space size represented by a single voxel point in the three-dimensional frequency domain coherent optical tomographic image and the actual size of the image; segmenting a retinal base Bruch's membrane from the fundus radian model by using a single-layer graph searching method; calculating the space tensor of each voxel point so as to obtain the point probability of the voxel point in a choroid vessel, and screening out a pre-selected area of the choroid vessel with the point probability value larger than zero; selecting out a point probability value larger than 75% and a point probability value less than 25% from the pre-selected area of the choroid vessel to serve as the high and low thresholds of the pre-selected area growth, so as to obtain the initial segmentation of the foreground and background of the initial segmentation of the choroid vessel, and solving the optimal model of a structure diagram according to the foreground and background of the choroid vessel to obtain the accurate segmentation of the choroid vessel.

Description

technical field [0001] The invention relates to a three-dimensional medical image quantitative analysis technology, in particular to a choroidal blood vessel segmentation method and system based on a three-dimensional coherent tomography image. Background technique [0002] Spectral-Domain Optical Coherence Tomography (SD-OCT) is an imaging technique that can produce three-dimensional micron-level precision images. Compared with traditional high-energy rays (X-ray, CT, etc.), OCT uses low-energy near-infrared light to provide non-invasive and non-invasive scanning imaging. It is the only technology that can realize three-dimensional human eye image reconstruction in vivo at this stage. During the imaging process, the detection light passes through the optic nerve fiber layer and the retinal pigment epithelial cell layer with high scattering properties before reaching the choroid, which leads to certain difficulties in OCT imaging of the choroid. In recent years, there have...

AI Technical Summary

Problems solved by technology

[0004] Existing quantitative analysis methods are either unable to analyze the choroidal structure, or can only perform incomplete analysis on the two-dimensional OCT interface, and they cannot accurately segment the three-dimensional choroidal vessels to calculate various clinical indicators based on the volume of the choroidal vessels

Images