Non-rigid multi-modality medical image registration method based on discrete optimization of ZMLD (Zernike Moments based Local Descriptor) and GC (Graph Cuts)

Abstract

The invention discloses a non-rigid multi-modality medical image registration method based on discrete optimization of ZMLDs (Zernike Moments based Local Descriptor) and GCs (Graph Cuts) and relates to the technical field of medical image processing. The method mainly comprises the steps of respectively calculating the ZMLD of a reference image I and a floating Image J, constructing an energy function by using an absolute error between the ZMLDs of the images I and J and an SAD (Sum of absolute differences) as data items of the energy function and a first-order derivative of a displacement vector field as a smooth item; solving a minimum value of the discretized energy function by using an alpha expansion optimization algorithm of the GC, and outputting an optimal displacement vector fieldcorresponding to the minimum value of the energy function, that is, a registrated image. According to the non-rigid multi-modality medical image registration method based on the discrete optimizationof the ZMLD and the GC, the problems that intensity and edge and textural features of an image cannot be accurately and simultaneously extracted, and the continuous optimization calculation is relatively high in complexity and is prone to local optimum in an existing method when noise and intensity distortion of the image occurs in a non-rigid image are solved. Experiments show that the precisionand efficiency of the non-rigid multi-modality medical image registration are improved by using the method of the invention.

Description

technical field [0001] The invention relates to the technical field of medical image processing, in particular to a non-rigid multimodal medical image registration method based on ZMLD and GC discrete optimization. Background technique [0002] In clinical medicine, different imaging modalities can provide different physiological information. The information provided by unimodal medical images is often limited. Multimodal medical image registration is conducive to complementing the information between different modal images. The information complementary images provide a variety of information about diseased tissues or organs, and provide a strong theoretical basis for doctors to make accurate diagnoses. [0003] At present, the calculation methods for the similarity measure in multimodal medical image registration are mainly divided into two categories. One is to use the information theory metric as the similarity measure. Mutual information (MI) is a widely used informat...

AI Technical Summary

Problems solved by technology

[0005] When there is noise and intensity distortion in a non-rigid image, the existing methods cannot accurately extract the mixed information of the image at the same time, and the continuous optimization has relatively high computational complexity and is easy to fall into the problem of local minimum

Images