Purpose: To evaluate image quality of coronary CT angiography with retrospectively ECG-gated 16 multi-slice spiral CT (MSCT), reconstructed with 0.75 mm slice thickness for optimal spatial resolution and with 1.3 mm slice thickness, to produce spatial resolution comparable to a 4-MSCT.
Materials and methods: Ten patients underwent coronary CT angiography with a 16-MSCT (Siemens Sensation 16, Forchheim, Germany) with 0.75 mm detector collimation. Raw helical CT data were retrospectively reconstructed using two different settings. Setting A: B20f smooth kernel, axial MPR with 1.3 mm slice thickness and 0.7 mm increment. Setting B: B35f "HeartView" medium-smooth kernel, 0.75 mm slice thickness, 0.5 mm increment. In the axial slices two regions of interest (ROIs) were placed in the area of the aortic root (AR) and more caudal in the area of the left ventricle (LV). Image noise was determined by the standard deviation of the CT numbers.Two readers determined visibility of coronary arteries by standardized maximum intensity projections (MIP) post-processing in left, right anterior and left anterior oblique projection plane from setting A and B. Each projection was rated on a five point rating scale concerning plaque delineation. Points determined for each data set were summed up and used for comparison.
Results: No significant difference between the CT-numbers was found for setting A and B (A: 283.0 in AR/295.9 in LV and B: 282,9 in AR/297.2 in LV; p >0.2). However, the image noise was significantly different for setting A and B (A: 4.46 in AR/1.67 in LV and B: 8.16 in AR/7.38 in LV; p <0.01). Better delineation of the coronary arteries and atherosclerotic lesions could be achieved from MIP projections in setting B compared to setting A.
Conclusion: Higher image noise is present in coronary 16-MSCT with thin-slice reconstruction compared to simulated 4-MSCT. However the MIP-reconstructions benefit most from the higher spatial resolution.