Computer-Aided Decision Support System for Catheter-Based Pulmonary Valve Replacement

点击翻译按钮获取中文摘要

ABSTRACT The aim of this proposal is to design, develop, and validate an interactive easy-to-use computer modeling tool that will help choose the best Transcatheter Pulmonary Valve Replacement (TPVR) device for an individual patient. Increased availability of a variety of self-expanding TPVR devices with varying shapes and sizes increases the number of candidates for TPVR but will require tools that can improve the efficiency and accuracy of determining patient eligibility as well as the optimal, patient-specific device. Current screening techniques are labor intensive, expensive, and continue to rely primarily on basic 2D morphological measures from CT reconstructions of the right ventricular outflow tract (RVOT) and implantation of actual devices into 3D printed models to visually assess device fit. Therefore, a clinical-grade application warrants integration of realistic physics of device mechanics, deformable vessel walls and their interaction at higher accuracy, as well as generation of quantitative metrics and visualizations within an intuitive user interface. The proposed tool will model the vessel wall, self-centering of the self-expanding device and their interactions using finite elements which will be incorporated as an external module for 3D Slicer-an open- source medical imaging platform. Metrics such as device dimensions, compression, maximum stress from contact, and assessment of proximal and distal circumferential seal will be automatically computed, visualized and graphed for the cardiologists, facilitating determination of patient candidacy and the optimal choice of a device for an individual patient. Our tool improves upon the current methods by incorporating (i) high-fidelity contact modeling including a frictional contact model between the device and the vessel wall, (ii) automatic estimation of the crucial metrics, (iii) near real-time performance allowing for incorporation into clinical workflows, and (iv) methodology verification and clinical validation. Modeling accuracy will be verified by comparing models based on existing pre-implant CT scans to post device implant CT scans in an ovine model of pulmonary insufficiency. Further, the 3D Slicer extension tool will be evaluated by three experienced interventional cardiologists (> 5 years of experience) at CHOP, who routinely perform TPVR implant for its (1) ease of navigation; (2) quantitative metrics; (3) visualizations, and (4) overall utility for clinical decision making.