Image quality and artifacts in coronary CT angiography with dual-source CT: initial clinical experience

We aimed to characterize artifacts observed in a routine clinical coronary CT angiography (CCTA) performed by a dual-source CT (DSCT) scanner (Definition; Siemens Medical Solutions).

Studies of 167 consecutive patients referred for CCTA, performed after beta-blockade (if not contraindicated), were prospectively analyzed for artifacts with a predefined visual approach. American Heart Association coronary segments (n = 2589) were assessed in 40%-80% R-R interval phases by 2 experts for stenosis, plaque presence or composition, and presence or type of artifacts. Each segment was considered evaluable when image quality was diagnostic in at least one cardiac phase. Artifacts included motion (cardiac, respiratory, patient), phase misregistration because of varying heart beats, calcified plaque blooming or beam hardening, metal beam hardening, large patient size, and contrast timing error.

Maximum HR (HR) during CCTA ranged from 45 to 120 beats/min (66.4 +/- 14.8 beats/min). Artifacts of some type were observed in 69 (41.3%) of 167 studies. Calcified plaque was the most common source of artifacts (14.4%), followed by misregistration (13.8%). Only 25 (1%) of 2589 coronary segments, in 6 (4%) of 167 patients were unevaluable, primarily because of calcified plaque blooming (coronary calcium score [CCS], 1112 +/- 1255]. Artifacts were associated with CCS (P = 0.002), change in HR (P = 0.01), age (P = 0.03), and body mass index (P = 0.048). The optimal phase for evaluation of all coronary arteries was 70% (mid-diastole), with a shift toward the systolic phases for HR > 70 beats/min.

CCTA artifacts with DSCT were related primarily to calcified plaque and cardiac phase misregistration. When correctly recognized, the artifacts did not have a serious effect on the final interpretation.