Progressive three-dimensional biomedical image registration method based on depth self-calibration

Abstract

The invention relates to a progressive three-dimensional biomedical image registration method based on depth self-calibration. Compared with the prior art, the problem of information leakage in a recursive cascade registration strategy and the defect that the registration precision is limited due to smooth constraint on a displacement field in a one-shot registration strategy are overcome. The method comprises the following steps: obtaining and preprocessing a three-dimensional biomedical image data set; constructing a registration model; training a registration model; obtaining and preprocessing an image to be registered and a template image; and obtaining a registration result of the three-dimensional biomedical image. According to the method, the previous registration result is input into the network in a recursive mode for self-calibration optimization, the displacement field generated by recursion each time is recorded and integrated, large displacement in the space is completed in a segmented mode through the recursive network, and therefore the registration precision of the three-dimensional biomedical image is improved on the premise of no information leakage, and GPU resources are effectively saved.

Description

technical field [0001] The invention relates to the technical field of three-dimensional biomedical images, in particular to a progressive three-dimensional biomedical image registration method based on depth self-calibration. Background technique [0002] With the development of medical imaging technology, modern medical diagnosis cannot be separated from medical image analysis. In order to analyze the condition of disease or changes in tissue structure, medical images captured from different locations and time periods should be aligned into a common space, resulting in medical image registration techniques, which aim at two or more image voxels Establish accurate non-linear anatomical correspondence between them. Therefore, it has been widely used in various clinical tasks, such as medical image fusion, treatment planning, and surgical navigation. [0003] In the past few decades, researchers have proposed many image registration methods, which can be roughly divided int...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to solve the problem of information leakage in the recursive cascade registration strategy and the defect of limited registration accuracy caused by the smooth constraint on the displacement field in the one-shot registration strategy, and to provide a depth self-calibration-based Progressive 3D biomedical image registration method to solve the above problems

Images