Registration method and device for CT image and MRI three-dimensional image

Abstract

The present disclosure provides a registration method and device for a CT image and an MRI three-dimensional image, wherein the method includes: acquiring the originally collected CT image and the MRI three-dimensional image and performing a preprocessing operation to obtain a CT processed image and an MRI processed image; The mode conversion generator obtained through training performs mode conversion on the CT processed image and the MRI processed image to obtain a CT mode converted image and an MRI mode converted image; CT processed images and MRI processed images are predicted to obtain single-modal registration confidence; using the pre-trained multi-modal registration network, the CT processed images, MRI processed images, CT modality conversion images, MRI modality Registration is performed on the state conversion image and the single-modal registration confidence level, so as to realize the automatic registration of the CT image and the MRI three-dimensional image.

Description

technical field [0001] The present disclosure relates to the technical field of digital medicine, and in particular, to a method and device for registering CT images and MRI three-dimensional images. Background technique [0002] Medical image registration is an important step in medical image processing. Traditional medical image registration methods include iterative closest point method, Gaussian mixture model method, congruent four-point set method, etc. These methods generally have the problem that the iterative optimization process is complicated in calculation, and it is easy to fall into local optimum and cause registration failure. In recent years, with the wide application of deep learning in the field of image processing, such traditional registration methods are gradually replaced by deep learning methods. [0003] Deep learning methods use convolutional neural networks to learn input image features and are widely used in the field of image processing. Among th...

AI Technical Summary

Problems solved by technology

However, current deep learning methods generally require a large amount of existing medical image data for network training. Considering that medical image data is difficult to obtain, and non-rigid registration deformation results between images are difficult to manually label, supervised deep learning registration methods are usually difficult. practical application

However, the unsupervised learning registration method needs to calculate the similarity loss function according to the appearance difference between the images to be registered. When the images to be registered have obvious differences in appearance in different modalities, it is difficult for existing methods to measure their similarity.

[0004] Therefore, such methods are difficult to apply to computer tomography (CT) and magnetic resonance (MRI) cross-modal image registration problems

Images