A Robust Registration Algorithm for Image-Guided Surgical Robot

IEEE Access

卷:6

Xin Ma¹  Qianqian Li¹  Rui Song¹  Xiaojing Liu² 

[1] Center for Robotics, School of Control Science and Engineering, Shandong University, Jinan, China;

[2] Department of Oral and Maxillofacial Surgery, School and Hospital of Stomatology, Peking University, Beijing, China;

关键词: Image-guided surgery;    image-guide surgical robot;    medical image registration;    point-based registration;    robust registration algorithm;   

DOI  :  10.1109/ACCESS.2018.2853601

来源: DOAJ

【 摘 要 】

The aim of this paper is to propose a robust registration algorithm to reduce the impact of the marker-recognition error on image-to-physical registration in the robot-assisted cranio and maxillofacial (CMF) surgery. Since the image-guided technology has been widely used in the CMF surgery, the surgical robot based on remote or cloud plan data has come into focus. As a critical procedure of the image-guided surgical robot, the image-to-physical registration has become a decisive factor of the operation result. The recognition error of the reference points is a main challenge for the registration. Therefore, we propose an improved method to cope with the noise in the image space. The image-to-physical registration is accomplished via a group of implanted reference markers. Firstly, the reference markers in the image space are picked out manually and extended to a fuzzy point set via a directional region growing algorithm. Then, the reference markers in the physical space are localized by the navigation cameras. At last, the transfer matrix is calculated using an improved registration algorithm based on the geometrical features of reference points in the two spaces. The experimental results demonstrate that the proposed method is less sensitive to the number of the reference points, and the influence of recognition noise can be decreased effectively. This paper is a foundational link of the cloud-based surgical robots, and the results of every case will be collected to the cloud for further data mining and analyzing.

【 授权许可】

Unknown