Biography
Enrollment Date: 2013
Graduation Date：2016
Degree：M.S.
Defense Date：2016.05.30
Advisors：Xiang Xie
Department：Institute of Microelectronics，Tsinghua University
Title of Dissertation/Thesis：Research on High Accuracy Surgical Navigation of Robotic Arm with Surgical Needle
Abstract:
In order to reduce major diseases incidence and mortality, a key work is early diagnosis and treatment. In traditional surgery, needle insertion is controlled by the surgeon depending on their own experience through CT image to obtain early biopsy sample. There are two problems in this manual insertion. One is the accuracy of the insertion is low and the probability of repeat surgery is high, the other is the repeat surgeries cause great harm to patients and may lead to complications. To solve the problems mentioned above, this thesis proposes a surgical navigation system with high accuracy optical measurement system and robotic arm collaborative control system to avoid the problems caused by manual intervention. In the designed system, by precisely controlling the robotic arm with surgical needle through binocular stereovision system, the surgical needle is moved to the objective insertion point and attitude which are planned by the surgeon. The achievements and innovations of this thesis are mainly in the following two aspects. One is the three-dimensional measurement of surgical needle’s position and attitude by binocular stereovision system, the other is the design and implementation of the surgical navigation system. The three-dimensional measurement of surgical needle’s position and attitude by binocular stereovision system is a vital technique in the designed surgical navigation system.
For the three-dimensional measurement of surgical needle’s position and attitude by binocular stereovision system, it includes two parts. One is the measurement of needle’s attitude, the other is the measurement of needle tip. For attitude measurement, needle axis extraction is a vital procedure. This thesis proposes a new algorithm, called “center point method”, for needle axis extraction. It has been proved, both theoretically and experimentally, to be more accurate than related method and the accuracy is promoted by nearly 70%. For needle tip measurement, the real needle tip in image is detected by a Harris detector. According to the experimental results, the measurement error of attitude and needle tip is 0.3494°±0.2455° and 0.76mm±0.28mm respectively, which meet the current demand for minimally invasive surgical precision.
For the design and implementation of the surgical navigation system, a preliminary framework of the new surgical navigation system is designed and we focus on the surgical navigation algorithm designed for three modules, including surgical navigation system calibration module, robotic arm control parameter calculation module and three-dimensional surgical needle tracking module in arithmetic processing unit. For surgical navigation system calibration module, a system calibration algorithm based on visible markers is proposed, which can ensure the sterility of surgical needle tip during the calibration procedure. The calibration errors of needle tip and needle’s attitude are 1.37mm±0.71mm and 0.62°±0.31° respectively, which meet the current demand for minimally invasive surgical precision. For robotic arm control parameter calculation module, combined with the calibration parameters obtained by surgical navigation system calibration module, an initial control algorithm is designed, which can ensure the single side markers in face of cameras after robotic arm’s initial movement according to the computed initial control parameters. An intuitive verification experiment is also designed to verify the target reachability of this initial control algorithm. For three-dimensional surgical needle tracking module, a locating and tracking algorithm is designed, which is the foundation of robotic arm trim control in the future.