A Miniature Navigation System for CT-guided Intervention

Computed Tomography (CT)-guided intervention has become an integral component of modern patient care. However, due to the lack of real-time visualization feedback, the workflow of the CT-guided intervention involves repeated CT scans, that often lead to prolong operation time and higher radiation dose to the patient. The objective of this dissertation is to propose a navigation system to streamline the current CT-guided intervention workflow. The goal is to develop a miniature patient-mount navigation system for spinal diseases. The miniature design is the key concept of the system to enable tracking capability without the lost of the accuracy. Another important feature is the auto-registration technique to eliminate the cumbersome registration procedure. To further improve the performance, an intuitive user interface utilizing the virtual bi-plane fluoroscopy technique is proposed. The integrated user interface allows surgeons to follow their already-mastered interventional technique without learning a new protocol. To evaluate the feasibility of the system, three validations are conducted. The first two validations, one with physical phantom and the other with in vitro porcine model, are used to validate machine accuracy and operation error, respectively. The third validation, involving an in vivo clinical human trial, is used to validate the clinical feasibility and efficacy of the system on percutaneous pulsed radiofrequency stimulation of dorsal root ganglion (PRF-DRG). The in vivo human trial shows machine position error 4.6 mm and tilting error 3.6°, which are within the range of conventional fluoro-guided technique