Automated Cardiac Phase Selection with 64-MDCT Coronary Angiography

Abstract

The aim of this study was to assess three different phase-selection methods for obtaining optimal CT coronary artery image quality. ECG-gated CT coronary angiography scans of 40 patients (23 men, 17 women; mean age, 56 years) were retrieved. The patient group was composed of 20 consecutive patients with heart rates < or = 65 beats per minute (bpm) and 20 consecutive patients with heart rates > 65 bpm. Three phase-selection methods were evaluated: fixed phase selection, manual phase selection, and automated phase selection. Two scoring systems were used to evaluate diagnostic quality: scoring of axial images on a 5-point scale and scoring of multiplanar reconstructions (MPRs) on a forced-choice 3-point preference scale. Differences were tested by Wilcoxon's signed rank test for the entire patient group and the two subgroups including patients with heart rates < or = 65 bpm and those with heart rates > 65 bpm. Axial image evaluation of the entire patient group showed statistically significant superior image quality for the manual phase-selection method compared with the predefined phase-selection method and no statistically significant differences were found for the other comparisons. Analysis at heart rates < or = 65 bpm showed no significant differences between phase-selection methods. Analysis at heart rates > 65 bpm showed the best results for the automated phase-selection method, and image quality was significantly better for the automated and manual phase-selection methods than for the predefined phase-selection method. The automated phase-selection method accurately detects the optimal diagnostic phase for CT coronary artery evaluation and has the potential to reduce operator time needed for image reconstruction.