Medical image fusion method based on 3D complex shear wave transform domain generalized statistics related model

Abstract

The invention discloses a medical image fusion method based on a 3D complex shear wave transform domain generalized statistics related model, wherein 3D sharing wave transformation of a complex domain is used as a sparsity expression tool. Compared with traditional wavelet transform, real number domain shear wave transform and the like, the medical image fusion method has advantages of high operation efficiency and high coefficient expression efficiency. For aiming at a defect that effect of relation between sub-band coefficients with different amounts in different directions are not considered in a traditional fusion method, a hybrid multi-scale related model is established, namely a real-number coefficient multi-scale multi-variable generalized Gaussian model and an imaginary-number part double-peak model; and the hybrid multi-scale related model is used in fusion. The beneficial effects of the medical image fusion method can be obtained according to description to the medical image fusion method. Compared with a traditional method, the medical image fusion method has advantages of overcoming a defect of insufficient sparsity expression in the traditional method, better extracting characteristic information in a to-be-fused image, and improving quality of a fused image.

Description

technical field [0001] The invention relates to the field of medical image fusion, in particular to a medical image fusion method based on a generalized statistical correlation model in a 3D complex shearlet transform domain. Background technique [0002] Multimodal medical image fusion plays an important role in clinical diagnosis. It is widely used in image-guided surgery, image-guided radiotherapy, non-invasive diagnosis and treatment planning. It is a key link in modern medical visualization technology. , widely used in modern medical clinical diagnosis. [0003] At present, multimodal medical image fusion methods can be roughly divided into three strategies: replacement methods, arithmetic methods, and multi-scale decomposition methods. Among them, the replacement method, such as the color space transformation method, may cause the distortion of the image spectrum; the arithmetic method, such as the Bayesian estimation method, may easily lead to a decrease in image con...

AI Technical Summary

Problems solved by technology

[0009] (2) Some literatures propose a fusion method based on translation-invariant shearlet transform. However, this translation-invariant property is realized by the non-subsampling Laplacian pyramid algorithm, and the operation time is long, which leads to the whole The fusion approach is very inefficient, which is not conducive to the promotion of clinical practice

[0010] 2. Another core issue of the multi-modal medical image fusion method based on multi-scale decomposition is the selection of fusion rules for low-frequency coefficients and high-frequency coefficients

As a result, it is easy to cause the fusion result to be very blurred, and there is a block effect, so that the visual effect of the entire fusion image is very poor.

Images