Stent evaluation in low-dose coronary CT angiography: effect of different iterative reconstruction settings.

BACKGROUND: Different iterative reconstruction (IR) techniques compensate increased noise from lower tube current-time product settings, yet the differences between IR settings remain unclear. OBJECTIVE: Aim of this study was to test whether different IR settings have a clinically relevant influence on image quality and on the diagnostic accuracy of low-dose CT angiography in patients with a stent. METHODS: Forty-two patients with 73 coronary stents were prospectively enrolled. Data were acquired with dual-source CT, and images were reconstructed with standard filtered back projection (FBP) and raw data-based IR with different settings (I3, I4, I5). Quantitative parameters, including CT-attenuation, noise, signal-to-noise ratio, contrast-to-noise ratio, as well as the presence of in-stent stenosis > 50% were determined. All patients had invasive angiography as reference standard. RESULTS: Mean effective dose was 0.32 ± 0.02 mSv. Image noise decreased significantly compared with FBP (I3 = 29%; I4 = 38% and I5 = 45%), whereas signal-to-noise and contrast-to-noise ratios increased significantly (all IR settings P < .01). Subjective image quality was superior with all IR settings (P < .01). FBP sensitivity, specificity, positive predictive value, and negative predictive value were 83%, 71%, 36%, and 96% per stent compared with 100%, 76%, 44%, and 100%, respectively, in IR reconstructions independent of the IR setting applied. CONCLUSION: In low-dose coronary CT angiography, higher IR settings significantly improved subjective and objective image quality but had no effect on accuracy.

Low-dose dual-source CT angiography with iterative reconstruction for coronary artery stent evaluation.

OBJECTIVES: The purpose of this study was to evaluate the image quality and diagnostic accuracy of very low-dose, dual-source computed tomography (DSCT) angiography for the evaluation of coronary stents. BACKGROUND: Iterative reconstruction (IR) leads to substantial reduction of image noise and hence permits the use of very low-dose data acquisition protocols in coronary computed tomography angiography. METHODS: Fifty symptomatic patients with 87 coronary stents (diameter 3.0 ± 0.4 mm) underwent coronary DSCT angiography (heart rate, 60 ± 6 beats/min; prospectively electrocardiography-triggered axial acquisition; 80 kV, 165 mA, 2 × 128 × 0.6-mm collimation; 60 ml of contrast at 6 ml/s) before invasive coronary angiography. DSCT images were reconstructed using both standard filtered back projection and a raw data-based IR algorithm (SAFIRE, Siemens Healthcare, Forchheim, Germany). Subjective image quality (4-point scale from 0 [nondiagnostic] to 3 [excellent image quality]), image noise, contrast-to-noise ratio as well as the presence of in-stent stenosis >50% were independently determined by 2 observers. RESULTS: The median dose-length product was 23.0 (22.0; 23.0) mGy · cm (median estimated effective dose of 0.32 [0.31; 0.32] mSv). IR led to significantly improved image quality compared with filtered back projection (image quality score, 1.8 ± 0.6 vs. 1.5 ± 0.5, p < 0.05; image noise, 70 Hounsfield units [62; 80 Hounsfield units] vs. 96 Hounsfield units [82; 113 Hounsfield units], p < 0.001; contrast-to-noise ratio, 11.0 [9.6; 12.4] vs. 8.0 [6.2; 9.3], p < 0.001). To detect significant coronary stenosis in filtered back projection reconstructions, the sensitivity, specificity, positive predictive value, and negative predictive value were 97% (32 of 33), 53% (9 of 17), 80% (32 of 40), and 90% (9 of 10) per patient, respectively; 89% (43 of 48), 79% (120 of 152), 57% (42 of 74), and 96% (121 of 126) per vessel, respectively; and 85% (12 of 14), 69% (51 of 73), 32% (11 of 34), and 96% (51 of 53) per stent, respectively. In reconstructions obtained by IR, the corresponding values were 100% (33 of 33), 65% (11 of 17), 85% (33 of 39), and 100% (11 of 11) per patient, respectively; 96% (46 of 48), 84% (129 of 152), 66% (47 of 71), and 98% (127 of 129) per vessel, respectively; and 100% (14 of 14), 75% (55 of 73), 44% (14 of 32), and 100% (55 of 55) per stent, respectively. These differences were not significant. CONCLUSIONS: In selected patients, prospectively electrocardiography-triggered image acquisition with 80-kV tube voltage and low current in combination with IR permits the evaluation of patients with implanted coronary artery stents with reasonable diagnostic accuracy at very low radiation exposure.

Image quality of ultra-low radiation exposure coronary CT angiography with an effective dose <0.1 mSv using high-pitch spiral acquisition and raw data-based iterative reconstruction.

OBJECTIVES: We evaluated the potential of prospectively ECG-triggered high-pitch spiral acquisition with low tube voltage and current in combination with iterative reconstruction to achieve coronary CT angiography with sufficient image quality at an effective dose below 0.1 mSv. METHODS: Contrast-enhanced coronary dual source CT angiography (2 × 128 × 0.6 mm, 80 kV, 50 mAs) in prospectively ECG-triggered high-pitch spiral acquisition mode was performed in 21 consecutive individuals (body weight <100 kg, heart rate ≤60/min). Images were reconstructed with raw data-based filtered back projection (FBP) and iterative reconstruction (IR). Image quality was assessed on a 4-point scale (1 = no artefacts, 4 = unevaluable). RESULTS: Mean effective dose was 0.06 ± 0.01 mSv. Image noise was significantly reduced in IR (128.9 ± 46.6 vs. 158.2 ± 44.7 HU). The mean image quality score was lower for IR (1.9 ± 1.1 vs. 2.2 ± 1.0, P < 0.0001). Of 292 coronary segments, 55 in FBP and 40 in IR (P = 0.12) were graded "unevaluable". In patients with a body weight ≤75 kg, both in FBP and in IR, the rates of fully evaluable segments were significantly higher in comparison to patients >75 kg. CONCLUSIONS: Coronary CT angiography with an estimated effective dose <0.1 mSv may provide sufficient image quality in selected patients through the combination of high-pitch spiral acquisition and raw data-based iterative reconstruction.