Unmarked medical image registration system and method in intraoperative navigation system

Abstract

The invention discloses an unmarked medical image registration system and a method in an intraoperative navigation system. The system comprises a three-dimensional scanner, a computer and a display device. The method comprises the following steps: through medical image data of a patient before an operation, a three-dimensional model is built, and a point cloud P is converted; a bone surface point cloud Q is acquired through the scanner in the operation; down sampling is carried out on two groups of point clouds respectively; a normal is calculated; feature calculation is carried out on the two groups of point clouds respectively; registration is carried out; a rotational translation matrix is obtained; the three-dimensional image before the operation is subjected to rotational translation; registration work is thus completed; and after initial registration, the later registration work only needs to use a nearest iterative point algorithm (ICP) for completion. No marked point needs to be used, and intraoperative unmarked accurate registration can be realized.

Description

technical field [0001] The invention relates to the field of medical images, in particular to a non-marker medical image registration system and method in an intraoperative navigation system. Background technique [0002] Intraoperative navigation as a medical navigation device refers to the accurate correspondence between the patient's preoperative and intraoperative patient anatomy, so that the doctor can clearly see the position of the surgical instrument and the patient's anatomy at a glance, making the operation more accurate. [0003] At present, intraoperative navigation generally has the following working principles. Doctors hold improved surgical tools and operate on patients’ surgical targets. The positional relationship between preoperative X-ray / CT, MR and other multi-mode images, after the corresponding coordinate transformation, controls the surgical tool to reach the required position, so as to implement the corresponding surgical operation. [0004] Among th...

AI Technical Summary

Problems solved by technology

[0004] Among them, determining the positional relationship between images is a very important part. At present, image registration is mainly divided into two methods: marked and non-marked. Using marked points makes registration faster, but in many cases, it is not convenient for patients. Adhesive markers, or surface adhesive points are easy to fall off, while implanted markers are more accurate but cause additional damage to the patient

[0005] The technical solution of the announcement number CN 106137395A also adopts the non-marker registration method, but uses the point cloud of the patient's surface, and the implementation of the method requires two sets of the same point cloud, which increases the difficulty of data acquisition

Images