Training method of image registration model, image registration method, equipment and medium

Abstract

The invention discloses a training method of an image registration model, an image registration method, equipment and a medium. The training method comprises the following steps: acquiring a to-be-registered training image and a reference training image corresponding to the to-be-registered training image; inputting the to-be-registered training image and the corresponding reference training image into a to-be-trained image registration model, and outputting a first training deformation field corresponding to the to-be-registered training image and a second training deformation field corresponding to the corresponding reference training image; calculating a first loss between an image obtained after the to-be-registered training image is registered by the first training deformation field and an image obtained after the corresponding reference training image is registered by the second training deformation field; and training an image registration model according to the first loss. According to the method, each pair of input images are registered to a certain state in the middle at the same time, the two deformation fields are output and combined into the total deformation field, the deformation between the to-be-registered images is reduced in a disguised mode, the registration difficulty between the images with large deformation is reduced, and the registration accuracy is improved.

Description

technical field [0001] The invention relates to the technical field of medical image processing, in particular to an image registration model training method, image registration method, equipment and media. Background technique [0002] Medical image registration refers to seeking one or a series of spatial transformations for a medical image to make it spatially consistent with the corresponding points of another medical image. This consistency means that the same anatomical point on the human body has the same spatial position on the two matching images. Medical image registration technology is the basis of medical image processing, and plays a very important role in the fields of image information fusion, auxiliary diagnosis, surgical planning, surgical navigation, and medical basic theory research. [0003] Among the traditional methods, the most representative non-rigid registration method for the elastic transformation of images is the Demons method. This algorithm is...

AI Technical Summary

Problems solved by technology

However, for most medical images, especially soft tissue organs, the deformation is large, and in most medical image registration applications, the registration requirements for soft tissue organs are relatively high. In this case, the Demons algorithm and most other The registration results of traditional registration algorithms are poor

Images