Summary of Previous Work

 

Endodontic Robot

 

        This project was conducted between 2009 and 2010 when I was in National Chiao Tung University (Taiwan). The purpose of the project was to improve the precision of the surgical movement in endodontic treatment. There were overall three subprojects, which were the design of the mounting mechanism, the robotic platform and the actuation system for drilling. I was responsible for developing the robotic platform, which would move and tilt the endodontic tools mounted on this platform. The CAD design of the robotic platform is shown as below:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

       

        The main difficulty of design is the requirement of dimension. Because the robot needs to be put in a patient's mouth, both the motors and the sensors need to be small. To fulfil the requirement, I deployed ultrasound motors (USM). The structure of a USM is shown in Fig. 1. A USM consists of a piezoelectric transducer, a driving rod, and a moving part. While the piezoelectric transducer vibrates above a particular velocity, the static friction between the moving part and the driving rod becomes kinetic friction. The static friction transfers the motion to the moving part whereas the kinetic friction allows motion between the moving part and the driving rod. This sequence allows a translational movement between the piezoelectric transducer and the moving part, which at high frequencies will appear as smooth motion. Based on this type of motor, a four degree-of-freedom (DOF) stage was devised.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

        In terms of position sensing, magneto-resistive (MR) sensors and magnetic rods were embedded in the robotic stage. Each MR sensor produces an output signal of two sine waves with a quarter cycle phase difference. An algorithm was applied to filter the non-linear segments of the output signal. The remaining filtered signal provides analogue information on the motor position. A laser Doppler vibrometer was used to calibrate the position data. Furthermore, a controller was also designed. To design the controller, the USM was first mathematically modelled. The model consisted of two parts, which were the model of a piezoelectric transducer and the model of friction. This USM model allowed me to simulate the motion of the motor. The simulation was later used for facilitating the controller design. Based on this USM model, fuzzy rules for the controller were designed in Matlab. In experiments, the fuzzy controller was applied to the USM stage via LabView and a real-time control board, National Instruments sbRIO-9631. The positioning error of step control was less than 0.072 mm. This high precision displayed the potential for the design to improve endodontic procedures. 

 

Rehabilitation Robot

 

        In early 2013, I joined the National Taiwan University as a research assistant. I was responsible for the mechanism of an eight DOFs upper-limb rehabilitation robot. Different from the mechanism of the endodontic robot, the mechanism of the rehabilitation robot was considerably larger and more complex. During my time as a research assistant, I greatly improved, through immersion, my ability to use computer-aided design software (Solidworks) and obtained practical experience using the belt driven system, harmonic drives, servo motors and potentiometers. One of the most significant design concerns of medical devices is the electromagnetic compatibility (EMC). Another project I was assigned, was to modify the configuration of another rehabilitation robot, made by the same research group, in order to pass the EMC examination. Through techniques of electrical grounding and shielding, this robot has achieved the requirements for EMC Accreditation (IEC 60601).

 

Concentric Tube Robot for Ophthalmic Surgery

 

        In September of 2013, I began my Masters course in Medical Robotics and Image Guided Intervention at the Hamlyn Centre, Imperial College London.  I was provided with the opportunity to devote myself to developing a concentric tube robotic (also known as an active cannula) system for vitreoretinal surgery. Throughout my masters, I have designed novel robotic mechanisms, modelled the kinematic model of the robot via Neural Network theory and performed experiments to evaluate the suitability of the device. This robotic system was designed for enhancing the dexterity of intraocular manipulation. In which, components such as Solidworks, 3D printer, Arduino and Labview have been intensively used to prototype mechanisms and drive actuators. The design was displayed below. On the left-hand side, the black component is a gripper to hold a nitinol tube. The gripper was mounted on a motion unit, which allowed the tube to be rotated and translated. The motion unit was designed as a modular unit (The detailed information can be found in DOI: 10.1109/EMBC.2015.7319583 ). 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Remote Sensing of Temperature during Magnetic Hyperthermia

        Since the Autumn of 2014, I have joined UCL Healthcare Biomagnetics Laboratory and started the project on magnetic hyperthermia. Hyperthermia is a technique of treating cancer by heating it. For thermal treatment like this, a reliable temperature sensing method is crucial, but the current gold standard, i.e., point measurement with thermal probes, is not reliable. Therefore, I proposed a novel method for sensing the temperature at a distance. The method is termed as magnetic particle thermometry, in which magnetic nanoparticles that used as heat source in hyperthermia are regarded as thermal probes. By analysing the signal of a field applicator, I am able to attain the average temperature of the particles, and thus the temperature of a tumour.  This concept has been presented at the 11th International Conference on the Scientific and Clinical Applications of Magnetic Carriers and will be the main focus of my PhD thesis. 

        The perspective of this project is different from the previous ones. This is less engineering-oriented and more science-oriented, which gave me an opportunity to immerse myself in the process of discovery, i.e., find a phenomenon, presume the reason behind the phenomenon, design experiments for hypothesis validation. Once the phenomenon has been fully realised, the phenomenon can then be applied to tackle issues in different scenarios. During the journey of discovery, Labview has been intensely used for data acquisition from multiple devices, e.g. digital oscilloscope, thermometer and Rogowski coil. In addition, I have attained another skill of using R to visualise and analyse experimental data, e.g., a script has been created for quantifying the heating power of magnetic nanoparticles (https://github.com/hoga85/ILP-Calculation).

 

NASA Lunar Sampling  System Design Challenge

        It was enjoyable to attend the challenge. With the constraints of volume, weight and power, it forced me to think if there is a way to achieve two functions with one single motion. After a few days of lacking sleep, a design was submitted (https://grabcad.com/library/spin-sampling-system-2). Due to license issue, I have picked up another CAD design software Inventor for this project.