A Real-Time Tracking and Visualization System for Robot-Assisted Template Location Method Applied to Lung Cancer Brachytherapy

Abstract

Accurate location of the puncture-guiding template, including the position and orientation, is essential for surgeons to implant radioactive seeds into an internal tumor in lung cancer brachytherapy. The objective of this paper is to establish a real-time tracking and visualization system (RTVS) to monitor the robot-assisted location process distantly and confirm the ultimate location accuracy without redundant computed tomography (CT) scans. RTVS consists of tracking and visualization components. A quaternion-based iterative closest point (QICP) algorithm for higher accuracy was proposed for the premised module of spatial registration. Arithmetic accuracy of QICP and clinical performance of RTVS were both validated by a series of experiments in a CT room. Spatial registration experiment shows that QICP consistently presents a distinctly higher degree of accuracy of 0.87±0.11 mm compared with other two conventional algorithms. RTVS, evaluated by tracking and visualization experiments, achieves a tracking accuracy of 1.05±0.05 mm position and (0.29±0.14) deg orientation. In addition, the time cost for template location is greatly reduced, so are the CT scan times. RTVS has the potential on lessening the workload of surgeons, reducing the CT radiation injury to the patient, and accelerating the progress of a brachytherapy surgery. The system presented means a new contribution to the lung cancer brachytherapy.