Abstract

Robot-assisted cardiovascular intervention (RCI) systems have shown success in reducing the x-ray exposure to surgeons and patients during cardiovascular interventional procedures. RCI systems typically are teleoperated systems with leader-follower architecture. With such system architecture, the surgeon is placed out of the x-ray exposure zone and uses a console to control the robot remotely. Despite its success in reducing x-ray exposure, clinicians have identified the lack of force feedback as to its main technological limitation that can lead to vascular perforation of the patient’s vessels and even their death. The objective of this thesis was to develop, verify, and validate mechatronics technology for real-time accurate and robust haptic feedback rendering for RCI systems. To attain the thesis objective, first, a thorough review of the state-of-the-art clinical requirements, modeling approaches and methods, and current knowledge gaps for the provision of force feedback for RCI systems was performed. Afterward, a real-time tip force estimation method based on image-based shape-sensing and learning-from-simulation was developed and validated. The learning-based model was fairly accurate but required a large database for training which was computationally expensive. Next, a new mechanistic model, i.e., finite arc method (FAM) for soft robots was proposed, formulated, solved, and validated that allowed for fast and accurate modeling of catheter deformation. With FAM, the required training database for the proposed learning-from-simulation method would be generated with high speed and accuracy. In the end, to robustly relay the estimated forces from real-time imaging from the follower robot to the leader haptic device, a novel impedance-based force feedback rendering modality was proposed and implemented on a representative teleoperated RCI system for experimental validation. The proposed method was compared with the classical direct force reflection method and showed enhanced stability, robustness, and accuracy in the presence of communication disruption. The results of this thesis showed that the performance of the proposed integrated force feedback rendering system was in fair compliance with the clinical requirements and had superior robustness compared to the classical direct force reflection method.