Surgical navigation space registration method based on improved PCA algorithm and DCP network

Abstract

The invention discloses a surgical navigation space registration method based on an improved PCA algorithm and a DCP network, and the method comprises the steps: 1), obtaining to-be-registered original point cloud data, including image space point cloud data and patient space point cloud data; 2) carrying out denoising and downsampling preprocessing operation on the point cloud data; 3) performing global registration based on an improved PCA algorithm on the preprocessed point cloud data so as to adjust the initial pose of the point cloud; and 4) taking the globally registered point clouds as the input of the DCP network, executing point cloud local fine registration, and applying the output registration relation of the DCP network to the two point clouds so as to realize registration of the two spaces. The method not only overcomes the defects of low registration precision, long consumed time, need of additional pasting of an external marking ball and the like in the existing space registration related algorithm, but also breaks through the limitation of the registration algorithm on the initial poses of the point clouds, that is, the method can meet all the initial poses of the point clouds in a surgical navigation use scene; and compared with a traditional space registration algorithm, the precision is higher, and the registration speed is higher.

Description

technical field [0001] The invention relates to the technical field of optical surgical navigation, in particular to a surgical navigation space registration method based on an improved PCA algorithm and a DCP network. Background technique [0002] The spatial registration based on optical surgical navigation is different from common medical image registration. It is the overall or local registration of two spatial point clouds, which is one of the applications of 3D point cloud registration technology. In neurosurgical procedures guided by optical surgical navigation systems such as puncture and ablation procedures, the spatial registration of the patient's head can be regarded as a rigid transformation, i.e. there is no shape transformation of the head, because the cranium confines the deformation of the brain , there are only changes in orientation, rotation, and position, translation. The patient space and 3D image space data are unified into the same coordinate system ...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art, and propose a surgical navigation space registration method based on the improved PCA algorithm and DCP network, which not only solves the problem of low registration accuracy and time-consuming problems in the current space registration related algorithms It takes a long time and requires additional pasting of external marker balls, etc., and it also breaks through the registration algorithm’s limitation on the initial pose of the point cloud, that is, the present invention can satisfy all the initial poses of the point cloud in the surgical navigation scene, and is better than the traditional The spatial registration algorithm has higher accuracy and faster registration speed

Images