Three-dimensional medical image segmentation method based on user interaction and shape prior knowledge

Abstract

The present invention discloses a three-dimensional medical image segmentation method based on user interaction and shape prior knowledge. The method comprises: finding out edge control points of an area to be processed in an image, calculating the outer normal vectors at the control points, employing the Hermit interpolation method to solve an interpolation coefficient, and obtaining an interpolation function; initializing a deformation model curved surface corresponding to the area to be processed into the image, overlapping the center of the model curved surface with the center of the area to be processed, taking the discrete vertex on the deformation model curved surface as an interpolated point, and employing the interpolating point and the interpolation coefficient to solve the level set function value of the interpolating point; employing the coordinates of the interpolating point and the interpolation coefficient to solve the level set gradient vector at each interpolating point; and performing curved surface iteration movement of the level set and the normal vector of the deformation model curved surface interpolating point to allow the edge of the deformation model curved surface to be gradually close to the edge of the area to be processed of the level set function being zero and generate a three-dimensional image.

Description

technical field [0001] The invention relates to a three-dimensional medical image segmentation method based on user interaction and shape prior knowledge, and particularly focuses on using user interaction information and shape prior knowledge to correct medical image segmentation. Background technique [0002] 3D medical image segmentation combines computer algorithms with prior knowledge of clinical anatomy and pathology to segment, extract and quantitatively analyze target organs and lesion areas, so as to obtain the 3D spatial structure or 3D functional feature distribution of the region of interest. Clinical applications such as disease diagnosis, treatment plan planning, and surgical navigation provide auxiliary information. [0003] Segmentation algorithms of medical images can be divided into automatic segmentation, manual segmentation and interactive segmentation. Automatic segmentation is a computer that completes all segmentation tasks independently. The whole pr...

AI Technical Summary

Problems solved by technology

Automatic segmentation is a computer that completes all segmentation tasks independently. The whole process does not require manual intervention, but the requirements for machines are very high, and due to the complexity and diversity of medical images, the results of automatic segmentation are generally difficult to meet medical requirements; It relies entirely on manual work, the entire segmentation process is very time-consuming, and requires very high experience and technical requirements for operators, which is not practical in actual operation.

Images