VIGILANT MEDICAL, INC. — Department of Health and Human Services SBIR Phase I: NHLBI

Description

Enhanced x ray angiography analysis and interpretation using deep learning OverMillion diagnostic X ray angiograms are performed annually in the US to guide treatment of coronary artery diseaseCADand cost over $billionDespite being the clinical standard of carevisual interpretation is prone to interand intra observer variabilityRecently as part of the NHLBI supported Prospective Multicenter Imaging Study for Evaluation of Chest PainPROMISEtrialour research team showed that cardiologists misinterpreted overof angiograms obstructive CADgreater thanvessel stenosisGiven the centrality of angiographic interpretation to the development of a treatment planreduced accuracy can lead to unnecessary poor outcomes and increased costs to our healthcare systemQuantitative coronary angiographyQCAoffers a computed measure of disease severityHoweverthe time and training required to perform QCA has largely relegated its use to research applicationsRecent advances in deep learning for medical image analysis offer an opportunity to address this problemThe potential impact is significant given increasing interpretation accuracy bycan positively benefit overpatients each yearThusour team proposes to develop an X ray angiographic analysis systemXAngiodriven by deep learning technology to enhance physician interpretationWe are uniquely positioned to accelerate development of XAngio by leveraging our team s PROMISE dataset of overangiograms with expert QCA scoringIn Phase Iwe will teach XAngio how to read angiographic images and how to discriminate obstructive CADXAngio will employ a convolutional neural networka computer vision technique rooted in deep learningto self characterize angiographic features from labeled imagesIn order to infer additional information from vast amounts of unlabeled imagesXAngio will incorporate a deconvolutional neural network technique to increase predictive performanceA pilot study of XAngio will test its ability to identify the presence of obstructive CAD in PROMISE angiogram imagesIf we are successfulwe envision a Phase II proposal which is focused on clinical translation of the technology and assessment of its impact on improving visual interpretation in a cohort of cardiologistsOur goal is to combine recent advances in deep learningbig data from PROMISEand scalable parallel computing to create XAngioIn the long termwe hope the combination of a cardiologist with XAngio as an assistive tool will improve the clinical accuracy of angiographic interpretation PROJECT NARRATIVE While invasive x ray angiography is the gold standard diagnostic tool for guiding immediate treatment in the cardiac cath labvisual interpretation of angiograms is challenging and subject to large interand intra observer biasesIn this projectwe will develop and validate XAngioa novel computer assistive technology to enhance cardiologist s angiographic reading and interpretation processIt is our hope that this technology will increase the accuracy of diagnosing obstructive coronary artery disease andultimatelyimprove patient outcomes