A method and device for extracting the centerline of a tubular target

Abstract

The present invention is applicable to the field of computer technology, and provides a method and device for extracting the centerline of a tubular target. The method includes: performing preprocessing on the image to be extracted for the centerline of the tubular target to obtain a corresponding enhanced image, and obtaining a gradient vector of the enhanced image Flow field, according to the gradient vector flow field, obtain the deformation force parameters corresponding to the initial ridge line segment list and extract the tubular target ridge point in the enhanced image, generate the initial ridge line segment list, and establish a regularized open curve deformation model. When the deformation end condition is not reached According to the model, a ridge segment in the initial ridge segment list is deformed and processed to obtain the corresponding tubular target center segment, and the initial ridge segment list is updated according to the ridge segments traversed during the deformation process. When the deformation end condition is reached, according to the tubular target The center line segment generates the center line of the tubular target, thereby optimizing the deformation process of the ridge line segment, reducing the deformation time of the ridge line segment, and improving the accuracy and extraction rate of the center line.

Description

technical field

[0001] The invention belongs to the technical field of computers, and in particular relates to a method and a device for extracting the centerline of a tubular target. Background technique

[0002] The quantitative extraction of centerline is of great significance for describing the characteristics of tubular objects and the development of virtual reality and augmented reality technology. Taking the computer-aided diagnosis and treatment of medical imaging as an example, the extraction of vascular centerline can effectively analyze the vascular structures and lesions of interest, then judge the luminal injury, establish the interventional operation path, and provide necessary help and guidance for the treatment of cardiovascular and cerebrovascular diseases.

[0003] There are many algorithms for extracting the centerline of tubular targets in noise images, and the commonly used methods are based on the shortest path, method based on topology refinement, meth...

Images