Search

Transcranial Magnetic Stimulation (TMS) is an excellent and non-invasive technique for studying the human brain. Accurate placement of the magnetic coil is required by this technique in order to induce a specific cortical activity. Currently, the coil is manually held in most of stimulation procedures, which does not achieve the precise clinical evaluation of the procedure. This thesis proposes a robotic TMS system to resolve these problems as a robot has excellent locating and holding capabilities. The proposed system can track in real-time the subject’s head position and simultaneously maintain a constant contact force between the coil and the subject’s head so that it does not need to be restrained and thus ensure the accuracy of the stimulation result. Requirements for the robotic TMS system are proposed initially base on analysis of a serial of TMS experiments on real subjects. Both hardware and software design are addressed according to these requirements in this thesis. An optical tracking system is used in the system for guiding and tracking the motion of the robot and inadvertent small movements of the subject’s head. Two methods of coordinate system registration are developed base on DH and Tsai-lenz’s method, and it is found that DH method has an improved accuracy (RMS error is 0.55mm). In addition, the contact force is controlled using a Force/Torque sensor; and a combined position and force tracking controller is applied in the system. This combined controller incorporates the position tracking and conventional gain scheduling force control algorithms to monitor both position and force in real-time. These algorithms are verified through a series of experiments. And it is found that the maximum position and force error are 3mm and 5N respectively when the subject moves at a speed of 20mm/s. Although the performance still needs to be improved to achieve a better system, the robotic system has shown the significant advantage compared with the manual TMS system. Keywords—Transcranial Magnetic Stimulation, Robot arm, Medical system, Calibration, Tracking

Three methods to improve the melt blown processing of postconsumer recycled poly(ethylene terephthalate) (PET) were explored: (1) using undried PET instead of dried PET; (2) blending recycled PET with other polymers; and (3) processing recycled PET at a higher temperature. The hydrolytic degradation occuring in the presence of moisture and heat, blending with PBT, blending with low I.V. PET, and the thermal degradation were expected to decrease the apparent viscosity of recycled PET during melt blown processing. Low viscosity contributes to production of quality webs, which have smaller fiber diameters, lower air permeability, and higher tenacity. The webs produced from undried PET, which had more moisture (0.263%) than dried PET (0.00713%), had lower intrinsic viscosity, smaller fiber diameters, lower air permeability and higher tenacity than those produced from dried PET. The load-elongation curves of webs produced from undried and dried PET were different. The undried PET webs had greater initial modulus, which indicated that the webs produced from undried PET had a more compact web structure. The recycled PET (I.V. = 0.82) was blended with either one of two grades of virgin PBT (I.V. = 0.88 and I.V. = 0.77) or a low molecular weight virgin PET (I.V. = 0.62). The webs produced from recycled PET blended with PBT had lower intrinsic viscosity, smaller fiber diameter, less air permeability and higher tenacity in the machine direction. The webs produced from recycled PET blended with low I.V. PBT showed greater property changes than those produced from recycled PET blended with high LV. PBT. On the other hand, the blending of low I.V. PET with recycled PET had lower intrinsic viscosity, smaller fiber diameters, less air permeability, higher filtration efficiency, greater tenacity, and exhibited greater shrinkage than 100% recycled PET. They were also thinner. It was also observed that the extent of change in the properties of the webs was proportional to the amount of low I.V. PET in the blend. Two different ways of using higher processing temperatures were explored: (1) the die temperature was increased from 295 to 315 °C; and (2) the temperatures of heating zones in the barrel were gradually increased from zone 1 to zone 4 in addition to increasing the die temperature to 315 °C. By only increasing the die temperature, no notable property improvements were achieved over those webs produced from recycled PET processed at normal temperatures. The webs produced by the later method had smaller fiber diameters, lower thickness and air permeability and greater tenacity in the machine direction. The effects of heat setting on the properties of melt blown webs were also studied to determine a more practical method to control the shrinkage of the melt blown PET webs. Through heat setting, shrinkage of webs was considerably reduced. The thickness was also reduced, resulting in more compact web structure. Tenacity in the machine direction increased when the webs were heat set at lower temperature for a short time. When heat setting temperature and time further increased, the tenacity in the machine direction decreased. The DSC results of webs before and after heat setting indicated an increase in the crystalUnity of fibers due to heat setting.