Sunday, December 1 | 3:20 p.m.-3:30 p.m. | S5-SSCA02-5 | E353C
In this Trainee Research Prize-winning presentation, researchers will review an AI model for coronary artery disease (CAD) risk stratification on CT myocardial perfusion imaging (MPI) and coronary computed tomography angiography (CCTA).
The model is designed for three purposes, according to presenter Yarong Yu of the Institute of Diagnostic and Interventional Radiology Shanghai Jiao Tong University Affiliated Sixth People's Hospital in Shanghai, China. It's intended to do the following:
- Fully automate quantification of myocardial blood flow (MBF) and ischemic myocardial volume.
- Diagnose hemodynamically significant stenosis with reference to invasive fractional flow reserve (FFR).
- Investigate the prognostic value of AI-MBF, particularly.
To train and evaluate the algorithm, Yu retrospectively gathered data from 268 patients who had underwent CT-MPI and CCTA. Yu divided the cohort randomly into an 8:2 ratio: a training dataset of 211 patients and a tuning dataset of 57 patients. Subsequently, 90 patients with stable angina were prospectively enrolled for external validation.
Each person initially underwent dynamic CT-MPI and then received invasive coronary angiography (ICA) or invasive FFR measurement within a one-month interval for the purpose of diagnostic performance assessment, with invasive FFR serving as the reference standard. Yu used multivariable Cox regression analysis to evaluate the prognostic value of AI-MBF for major adverse cardiovascular events (MACEs) in 660 patients from three distinct hospitals, using multivariable Cox regression analysis.
The AI-derived ischemic myocardial volume percentage improved risk stratification of CAD, according to the research. Among the major findings, the researcher noted excellent agreement between the AI model and manual measurements of both segment and patient-based MBF, for example.
Ninety participants with 116 target vessels were included for final analysis, which showed that the AI-derived ischemic myocardial volume enabled better risk stratification for CAD patients, the research noted. Visit the session for insight into the findings.