Aim: To investigate the potential dose reduction in cardiovascular computed tomography angiography (CTA) in a swine model using 320-detector volume CT with adaptive iterative dose reduction in three dimensions (AIDR 3D) reconstruction to maintain a comparable image quality (IQ) to that reconstructed by a conventional filtered back projection (FBP) algorithm.
Methods and materials: Twenty-four mini-pigs underwent cardiovascular CTA four times at 80 KVp and different tube currents. An automatic exposure control (AEC) system was used and the noise index (NI) was predetermined at a standard deviation (SD) of 20 (Method A, routine dose), and 25, 30, 35 (Methods B-D) to reduce the dose gradually. Method A was reconstructed with FBP. Methods B-D were reconstructed using AIDR 3D (strong). Two radiologists graded IQ by reviewing both cardiac and vascular structures using a five-point scale. Quantitative IQ parameters of image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were measured and compared. A receiver-operating characteristic (ROC) analysis was performed to select a radiation reduction threshold and maintain comparable IQ (score ≥4).
Results: Method B and C had significantly lower image noise (p<0.0001), higher CNR and SNR than Method A (p<0.0001). Compared with Method A (noise: 52.7±8.3; SNR: 11.7±2.8; and CNR: 9.9±2.7), Method C had comparable subjective IQ and higher objective IQ (noise: 38.9±6.1; SNR: 16.3±3.5; and CNR: 13.5±3.3). The results of the ROC curve showed that Method C (SD30) was the optimal dose threshold to maintain a comparable subjective IQ (AUC: 0.85, 95% confidence interval [CI]: 0.80-0.90). The effective dose (ED) of Method C was reduced by 49%, compared to that of Method A (0.33±0.08 mSv versus 0.65±0.15 mSv).
Conclusion: AIDR 3D at a strong level combined with an AEC system can potentially reduce the ED by 49% and maintain an IQ comparable to that achieved using a routine-dose and FBP reconstruction in mini-pig cardiovascular CTA.
Copyright © 2016 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.