Computer vision-based evaluation of prosthetic heart valves

点击翻译按钮获取中文摘要

We propose a project in collaboration with ViVitro Labs Inc., a Canadian company that offers industry-leading cardiovascular test equipment and related laboratory testing and consulting services. Through their cardiovascular testing and consulting services, ViVitro Labs Inc. helps to improve products meet regulatory requirements for independent third-party testing for FDA audit compliance and other regulatory submissions. We aim to address the issue of automating the analysis of video footage acquired during valve testing. Our investigations will focus on researching computer vision methods that can be used by the company for the automatic quantitative analysis of pulsatile and durability testing videos. The general idea is to develop quantitative evaluation metrics and apply them the testing videos; these metrics would complement current measurement techniques on pressure and flow data and replace the current qualitative, operator-based analysis of video footage. The development of a framework for the automatic quantitative analysis of pulsatile and durability testing videos would lay the ground work toward an automated production line tester that would effectively and reliably make a pass/fail assessment of the device performance when compared to a baseline.