Method and system for registering three-dimensional medical images on basis of weighted fuzzy mutual information

Abstract

The invention provides a method and a system for registering three-dimensional medical images on the basis of weighted fuzzy mutual information, and relates to the technical field of image-guided radiotherapy and medical image analysis. The method mainly includes steps of 1, guiding the medical images; 2, displaying the medical images; 3, processing the medical images; and 4, registering the medical images. In the step of guiding the medical images, single-mode image registration and multi-mode image registration are supported; in the step of displaying the medical images, cross section, coronal plane and sagittal plane images of the medical images to be registered are colored differently by a pseudo-color technology; in the step of processing the medical images, grayscales of the medical images are classified and compressed on the basis of a concept of fuzzy entropy, and mutual information calculation is reduced; and in the step of registering the medical images, normalized mutual information measures are modified on the basis of the fuzzy entropy, and the robustness of medical image registration is improved. The method and the system have the advantages that the measure method on the basis of the mutual information is adopted, the method and the system are applicable to single-mode image registration, and a good effect can also be realized for multi-mode image registration.

Description

technical field [0001] The invention relates to the technical field of image-guided radiotherapy and medical image analysis, in particular to a three-dimensional medical image registration method based on weighted fuzzy mutual information, which is applied to image-guided radiotherapy to determine the positioning error of a patient. Background technique [0002] The goal of radiotherapy is to maximize the concentration of radiation into the target area, so that the normal tissues and organs around the tumor are less or not irradiated, so as to kill tumor cells to the greatest extent and better protect normal tissues. Precision radiation therapy is an important means to achieve this goal. However, some uncertain factors in the radiation therapy process affect the precise radiation therapy. Among them, setup error is an important factor affecting the accuracy of radiotherapy, even for well-fixed head and neck tumors. In order to solve these problems, the radiotherapy machine...

AI Technical Summary

Problems solved by technology

At present, most of the automatic registration methods in the system are based on the registration of bone features, and its accuracy depends on the accuracy of bone shape feature extraction; however, due to factors such as organ movement, deformation, and treatment process, the position of the tumor relative to the bone changes; and bone-free areas (inside the pelvis, e.g., cavities) cannot use bone-based registration methods

The registration method based on voxel gray level does not need to extract features, and the registration method based on mutual information has been studied a lot; however, the registration method based on mutual information is also too sensitive to changes in gray information and lacks image data. It is more sensitive, has a large amount of calculation data and robustness, especially for 3D / 3D image registration, which has a large amount of data and a long calculation time. It is still not suitable for clinical application

Images