Wednesday, December 4 | 8:20 a.m.-8:30 a.m. | W1-SSNPM03-3 | S402
For this scientific session, Harvard researchers applied an AI-based body composition algorithm to 1,277 emergency abdominal CT exams from people in their 30s for the purpose of predicting major adverse cardiovascular events (MACE) and dyslipidemia.
Using a single-click image-processing AI tool called ClariMetabo (Claripi.ai) for inferencing all deidentified CT examinations performed on multivendor CT equipment, researchers obtained separate measures of CT attenuation values and volumes of subcutaneous and visceral fat, psoas, and all abdominal wall muscles, at T12-L4 vertebral levels.
Presenter Emiliano Garza Frias, MD, from the department of radiology at Massachusetts General Hospital and Harvard Medical School, will review how the AI demonstrated that abdominal visceral and waist circumference at multiple levels were the best and most significant predictors of MACE and dyslipidemia.
Specifically, abdominal visceral fat (hazard ratio [HR]:1.7; p < 0.02) and waist circumference (HR: 2; p < 0.03) predict higher cardiovascular disease-related MACE and dyslipidemia. Conversely, the negative HR as noted for absolute CT numbers (from -1.1 to - 1.6) likely offered protection from dyslipidemia due to less fat volumes, according to study details.
Attend the session to hear more about an example of opportunistic screening in young patients. It might be one way to increase the lead time for preventive steps to mitigate MACE and dyslipidemia, according to the researchers.