Cardiac CT and ultrasound image registration method based on salient region area matching

Abstract

The invention discloses a cardiac CT and ultrasound image registration method based on salient region area matching, and the method mainly solves a problem of low registration precision and low speed in the prior art. The method comprises the steps: 1) pre-processed CT and ultrasound images respectively serves as a reference image and a floating image, and a gray-scale feature pyramid model and a neighbourhood mean difference feature pyramid model of the two images are respectively established; 2) saliency maps of the reference image and the floating image are respectively generated based on above two pyramid models, and the saliency maps are binarized; 3) for the binarized results, regions of interest are extracted according to region area features, coarse registration based on region centroid distance and precise registration based on ICP algorithm are performed on the regions of interest, and registration parameters are obtained; 5) according to the registration parameters, rotation and translation transformation is performed on the floating image, and a registration result is obtained. According to the invention, the method has high registration precision and fast speed and can be used for real-time registration of preoperative CT image and intraoperative ultrasound image.

Description

technical field [0001] The invention belongs to the technical field of medical image processing, in particular to an image registration method, which can be used for real-time registration of preoperative CT images and intraoperative ultrasound images in heart surgery. Background technique [0002] With the continuous acceleration of the pace of life and the increasing pressure of life in modern society, the incidence of cardiovascular diseases is increasing year by year. According to the survey data of the World Health Organization, the number of deaths from cardiovascular diseases accounts for 30% of the total number of deaths in the world every year. It is estimated that by 2030, the number of deaths from cardiovascular diseases will increase to 23.3 million, and cardiovascular diseases will continue to be the number one cause of death threatening human health. Therefore, research on the diagnosis of cardiovascular diseases has received extensive attention, and more accu...

AI Technical Summary

Problems solved by technology

However, due to the poor quality of ultrasound images and the large difference in grayscale between CT images, the single grayscale-based registration accuracy is not ideal and takes a long time, which cannot meet the image registration accuracy and speed requirements of image-guided surgery.

The feature-based registration method converts the analysis of the entire image into the analysis of image features, which greatly reduces the amount of calculation in the image processing process and has better real-time performance, but the registration accuracy mainly depends on the geometry. Or the selection and extraction of anatomical features, the accuracy of feature extraction will have a great impact on the entire registration process, because the commonly used feature extraction mainly relies on manual selection, the speed is slow and human bias will be introduced, making the different modes of extraction The lower image features are quite different

Images