P016] Performance of 3 CT scanners: Results from a 4 year period

Abstract

Purpose Present the results of a 4 years period quality control routine test performed on 3 CT scanners (Siemens Somaton Sensation Open (S), Toshiba Aquilion 16 (T) and Ge HiSpeed Lx/i (G)) to assure compliance of the monitored parameters with baseline values and detect system performance malfunctioning. The weekly test, is a 15 minute check acquisition of the CATPHAN 600 phantom sections. Methods Baseline values were established using the average of 10 images in sequential mode 120 kV, 200 mAs(T), 280 mAs(S) and 300 mAs(G), collimation 1 × 5 mm(SG 31 mGy(G); 35 mGy(S). Performance indicators: Noise Index (standard deviation of the HU measured in a central ROI of 28 cm[b] (40% useful radius) Homogeneity index (maximum difference on the average HU measured in 4 periphery ROI’s relative to a central ROI each with 10cm2); Spatial Resolution Index (zoomed visualization of the last perceptible group of a lp/cm pattern using constant WW, WL); Low Contrast Index (CD – diameter * percentage contrast (mm*%) for each target – 3 groups of different sized low contrast targets 10HU, 5HU, 3HU – 2 images averaged; HU index (average HU in 45 mm[b] ROI’s of 8 target materials from −1000 up to 1000 HU. Results Noise index: σ(T) = 5.36 ± 2%; σ(G) = 2.87 ± 3%; σ(T) = 4.01 ± 5%; maximum deviations: 8.6%(T); 3.1%(G); 4.7%(S); outliers(>10%): two (12.3;16.0%(T)); one 21.6%(G); zero(S). Homogeneity index: HU = 9.53 ± 4%(T); HU = 7.1 ± 3%(G); HU = 12.7 ± 3%(S); maximum difference deviations: 14.7%(T); 14.5%(G); 12.6%(S); outliers(>15%): six(16.7;15.7%(T)) one 24.2%(G); zero(S). Spatial Resolution index (lp/cm): 9(T); 7(G); 8(S); CD index (mm*%): Conclusions The customized test proved to be a simplified but incomplete method for system performance consistency evaluation since it was inefficient in the detection of ring artifacts caused by faults in detector system calibration for the head protocol. The addition of a spiral acquisition in the homogenous section, lead to direct visualization and awareness of this type of artifacts but still do not prevent the detection of artifacts in larger FOVs or in special collimation sets depending on the CT scanner design.