Biography
Enrollment Date: 2015
Graduation Date：2018
Degree：M.S.
Defense Date：2018.05.24
Advisors：Xiang Xie
Department：Institute of Microelectronics，Tsinghua University
Title of Dissertation/Thesis：An Optical Tracker Based Robot Registration Method for Thoracentesis
Abstract:
The key work to reduce the morbidity and mortality of major diseases is early diagnosis and treatment. Currently, CT-guided percutaneous biopsy surgery is an important diagnostic method, which highly depends on the doctor's experiments to insert the needle. And then the result is confirmed by medical images. However, this surgery is high risk and low precision, which leads to a long operation time and causes complications. Therefore, Robot-assisted needle insertion technology is now widely used in minimally invasive surgeries. The success of these surgeries highly depends on the accuracy of the position and orientation estimation of the needle tip, that is, the accuracy of the registration method among multiple systems. The difficulty of registration between multiple systems is that: 1. The registration method is accurate and automated for the overall system. 2. Avoiding the influence of the dynamic error of the robot. 3. Verifying the accuracy and robustness of the proposed registration algorithm conveniently.
To the difficulty 1, we propose an optical tracker based automated registration method among surgical navigation systems, which can accurately calculate the multi-system transform relationship among the surgical needle, the CT device, the robot, markers and the optical tracker. Before surgery, we use CT images to the calculate the transform between needle and marker. And then, a three-step registration method is proposed to solve the transform among the robot, markers and the optical tracker, which include:1. Estimation of the rotational transform between the robot and the optical tracker. 2. Move the marker to the target point using feedback translational motion. 3. Register the robot and the optical tracker with the multiple pose of robot. In surgery, to the difficulty 2, a feedback control method is proposed to control the robot accurately place the surgical needle into the target position and orientation, avoiding the impact of the error of robot parameters.
To the difficulty 3, we designed a verification tool to accurately measure the needle placement position error and orientation error, avoiding the use of medical imaging to generate radiation. The verification tool has the multi-layer structure with different size of holes to verify the accuracy of different puncture directions. The needle will pass through a certain layer only if it meets some certain restriction of orientation and position. The more layers the needle passes, the tighter the restriction will be. After a large number of experiments, the result shows that the position accuracy of needle placement is below 0.4mm and the needle orientation accuracy is below 0.1 degrees. And theoretically the needle can hit the 1mm diameter tumor with the puncture depth less than 150mm. The result meets the current requirements of percutaneous puncture surgery. In addition, we design a simulated respiration model which can simulate the chest surface movement and tumor movement when patient in free breath. This work can be used in the future for the lung cancer.