The effect of heart rate on coronary plaque measurements in 320-row coronary CT angiography

Repeatability of quantitative assessment of atherosclerotic plaques is important for the accurate detection of high-risk plaques in coronary CT angiography (CTA). We assessed the effect of heart rate (HR) on plaque CT number using a coronary artery model and a cardiac phantom capable of simulating cardiac motion. The coronary artery model with luminal stenosis on a cardiac phantom was imaged with a simulated HR of 0, 50, 60, and 70 beats per minute using a 320-row CT scanner. We reconstructed CT images for cardiac diastolic phases (for 75% R-R interval) using filtered back projection (FBP), hybrid iterative reconstruction (AIDR3D), and model-based iterative reconstruction (FIRST). Two observers measured plaque attenuation in the lesion with 75% stenosis. The coefficient of determination (R-2) was obtained to evaluate interobserver agreement. At HR 70, FIRST improved the correlation between two observers compared with FBP and AIDR3D (FIRST: R-2=0.68, p<0.05; FBP: R-2=0.29, p=0.31; AIDR3D: R-2=0.22, p=0.18). These R-2 at HR 70 were lower compared with at HR 50 (FIRST: R-2=0.92, p<0.05; FBP: R-2=0.83, p<0.05; AIDR3D: R-2=0.87, p<0.05) and HR 0 (FIRST: R-2=0.97, p<0.05; FBP: R-2=0.89, p<0.05; AIDR3D: R-2=0.95, p<0.05). Higher HR affected plaque measurement repeatability in coronary CTA. FIRST may improve plaque measurement repeatability at the higher HR compared with FBP and AIDR3D.