• computed tomography
• coronary artery disease
• CT coronary angiography
• image processing
• radiation dose

INTRODUCTION AND PURPOSE. Computed tomography coronary angiography (CTCA) has seen a dramatic evolution in the last decade owing to the availability of multislice CT scanners with 64 detector rows and beyond. However, this evolution has been paralleled by an increase in radiation dose to patients, that can reach extremely high levels (>20mSv) when retrospective ECG-gating techniques are used. On CT angiography, reduction of tube voltage allows to cut radiation dose with improved contrast resolution due to the lower energy of the X-ray beam and increased photoelectric effect. Our purpose is twofold: 1) to evaluate the radiation dose of CTCA studies carried out using a tube voltage of 100kV and a low noise reconstruction filter, compared with a conventional tube voltage of 120kV and a standard reconstruction kernel; 2) to assess the impact of the 100kV acquisition technique on the diagnostic quality of 2D and 3D image reconstructions performed with open source software (OsiriX).
MATERIALS AND METHODS. Fifty-one non-obese patients underwent CTCA on a 64-row CT scanner. Out of them, 28 were imaged using a tube voltage of 100kV and a low noise reconstruction filter, while in the remaining 23 patients a tube voltage of 120kV and a standard reconstruction kernel were selected. All CTCA datasets were exported via PACS to a Macintosh™ computer (iMac™) running OsiriX 4.0 (64-bit version), and Maximum Intensity Projection (MIP), Curved Planar Reformation (CPR), and Volume Rendering (VR) views of each coronary artery were generated using a dedicated plug-in (CMIV CTA; Linköping University, Sweden). Diagnostic quality of MIP, CPR, and VR reconstructions was assessed visually by two radiologists with experience in cardiac CT using a three-point score (1=poor, 2=good, 3=excellent). Signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), intravascular CT density, and effective dose for each group were also calculated.
RESULTS. Image quality of VR views was significantly better with the 100kV than with the 120kV protocol (2.77±0.43 vs 2.21±0.85, p=0.0332), while that of MIP and CPR reconstructions was comparable (2.59±0.50 vs 2.32±0.75, p=0.3271, and 2.68±0.48 vs 2.32±0.67, p=0.1118, respectively). SNR and CNR were comparable between the two protocols (16.42±4.64 vs 14.78±2.57, p=0.2502, and 13.43±3.77 vs 12.08±2.10, p=0.2486, respectively), but in the 100kV group aortic root density was higher (655.9±127.2 HU vs 517.2±69.7 HU, p=0.0016) and correlated with VR image quality (rs=0.5409, p=0.0025). Effective dose was significantly lower with the 100kV than with the 120kV protocol (7.43±2.69 mSv vs 18.83±3.60 mSv, p<0.0001).
CONCLUSIONS. Compared with a standard tube voltage of 120kV, usage of 100kV and a low noise filter leads to a significant reduction of radiation dose with equivalent and higher diagnostic quality of 2D and 3D reconstructions, respectively in non-obese patients.