A study published in the Journal of the American College of Cardiology (JACC) Cardiovascular Imaging reports that researchers found that AI algorithms can more rapidly and objectively determine coronary calcium score (CAC) in CT and PET images than physicians. The model was trained using 9,543 CT scans from a cohort of 4,331 patients with major adverse cardiac events (MACEs) who underwent PET/CT imaging. MACE risk stratification in four CAC score categories (0, 1-100, 101-400 and over 400). The AI also performed well when studies were obtained from very-low-attenuation scans. The authors concluded that AI CT scores predict risk similarly to CAC scores obtained by experienced operators from ECG gated CAC scans.
