Abstract 12003: Effects of Iterative Reconstruction on Coronary Artery Calcium Scoring

Abstract

Purpose: IR is now commonly used in coronary computerized tomography to facilitate lower kV and mAS, decrease noise and improve resolution but has not been applied to coronary artery calcium scanning (CACS). This study was designed to evaluate the effects of varying levels of iterative reconstruction (IR) on CACS.

Methods: A calibrated anthropomorphic phantom with 7 calcium inserts of varying length and density was scanned using a Philips iCT scanner and Agatson Unit (AU) scoring was performed using standard software. One thousand seven hundred measurements were obtained at 80, 100 and 120 kV, from 10-100 mAs in 10 mAs increments, and at IR levels of 0-7. All data were analyzed independently by 2 separate observers.

Results: At standard CACS 120 kV/80 mAs settings without IR, the insert AUs of 2.8, 13.9, 41.5, 71.9, 73.9, 161.5 and 313.1 encompassed the range of clinically encountered values. At every combination of kV and mAs there were small but nonsignificant changes in CACS with increasing IR levels (Table). There were no changes in assignment to the standard clinical risk classification groups of 0, 1–10, 11–100, 101–400 and >400 AU with any level of IR. The interobserver variability was 0 (Kappa= 1.0). Table. AU of Phantom Inserts at Representative kV/mAs and Varying IR

Conclusions: 1) In a phantom, increasing IR levels had no significant effect on CACS and risk assignment over the full range of kV/mAs settings. 2) The potentially beneficial effects of IR on image appearance and resolution may be pursued without loss of accuracy.