Cardiovascular risk prediction from AI analysis of coronary calcifications

Cardiovascular risk prediction from AI analysis of coronary calcifications Summary To improve cardiovascular risk determination, we will create a software solution to predict future cardiovascular events from screening CT calcium score (CTCS) images using artificial intelligence and image analytics. Our group includes engineers and clinicians from PulseImaging.AI, Case Western Reserve University (CWRU), and clinical institutions at the forefront of CT calcium score screening: Houston Methodist and University Hospitals of Cleveland (UH). These medical institutions are promoting low-cost/no-cost CTCS exam programs, respec- tively, because they are convinced of the value to healthcare, collecting >170,000 CT calcium score exams over the past 10 years. Our AI methods greatly improve risk prediction as compared to the standard Agatston score, which has been identified as the best single predictor of future adverse events. Agatston basically sums up the amount of calcification in the heart, albeit in a non-linear way. Our approach builds upon pathophysiologic ob- servations and analyzes many features of calcifications (e.g., density, numbers of calcifications, number of ves- sel territories, and spatial distribution), a collection we call calcium-omics. With improved risk prediction, cardiol- ogists will be able to tailor personalized treatments for patients and effectively counsel them as to adherence to drug regimens and lifestyle changes. In Phase I, we will address important feasibility issues (e.g., automated robust calcium-omics feature assessments and software demonstrations) crucial for planning our product. Project Number: 1R41HL180169-01 | Fiscal Year: 2025 | NIH Institute/Center: National Heart Lung and Blood Institute (NHLBI) | Principal Investigator: Juhwan Lee (+1 co-PI) | Institution: PULSEIMAGING.AI, LLC, Cleveland, OH | Award Amount: $304,415 | Activity Code: R41 | Study Section: Special Emphasis Panel[ZRG1 ISB-Z (13)] View on NIH RePORTER: https://reporter.nih.gov/project-details/1R41HL18016901