Non-rigid registration method of prostate multimodal images based on mixture model

Abstract

The invention discloses a prostate multi-modal image non-rigid registration method based on a mixed model, and mainly solves the problem in the prior art that registration of prostate CT and MRI images is not good. The method is realized by steps of 1) inputting prostate CT and MRI images and segmenting the prostate CT and MRI images to obtain targets respectively; 2) taking filtered prostate CT and MRI images as a reference image and a floating image respectively, and subjecting the two images to pyramid multi resolving; 3) subjecting the floating image to offset and layer-by-layer iteration to obtain a crude registration image; 4) dividing the reference image and the crude registration image in a control point mesh manner, moving control points of the crude registration image, and obtaining a post-transformation image according to displacement of the control points; 5) continuously carrying out iteration until the similarity between the post-transformation image and the reference image is maximal, so as to obtain final displacement; and 6) obtaining a registration image according to the final displacement. The invention improves the registration precision between complex deformation and a multi-mode image, and can be used for medical CT and MRI image processing.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a non-rigid registration method for prostate images, which can be used for processing medical CT images and nuclear magnetic resonance MRI images. Background technique [0002] The prostate is the largest accessory gonad in men and one of the exocrine glands in the human body, located between the bladder and the original genital diaphragm. Genetic mutations in cells can cause uncontrolled proliferation and cause cancer. Prostate cancer is a malignant tumor from the prostate gland. In addition to expanding in size and invading adjacent organs, malignant cells may also metastasize to other parts of the body, especially bones and lymph nodes. Prostate cancer can cause pain, difficulty urinating, erectile dysfunction and other symptoms, seriously endangering people's health. In Western countries, prostate cancer is the second most common cancer in men, killing ...

AI Technical Summary

Problems solved by technology

The Demons algorithm based on the optical flow field model uses the gray gradient information of the reference image to determine the movement of each pixel of the floating image, but this method of only relying on the gradient information is often insufficient, especially when the gradient is very small. The direction of image deformation cannot be determined

Therefore, when the image has a large deformation, it cannot achieve better results

Images