The aim of this study was to demonstrate that dose reduction and constant image quality can be achieved by adjusting X-ray dose to patient size. To establish the relation between patient size, image quality and dose we scanned 19 patients with reduced dose. Image noise was measured. Four radiologists scored image quality subjectively, whereby a higher score meant less image quality. A reference patient diameter was determined for which the dose was just sufficient. Then 22 patients were scanned with the X-ray dose adjusted to their size. Again, image noise was measured and subjective image quality was scored. The dose reduction compared with the standard protocol was calculated. In the first group the measured noise was correlated to the patient diameter (rho=0.78). This correlation is lost in the second group (rho=-0.13). The correlation between patient diameter and subjective image quality scores changes from rho=0.60 (group 1) to rho=-0.69 (group 2). Compared with the standard protocol, the dose was reduced (mean 28%, range 0-76%) in 19 of 22 patients (86%). Dose reduction and constant noise can be achieved when the X-ray dose is adjusted to the patient diameter. With constant image noise the subjective image quality increases with larger patients.

This qualitative study is intended to create awareness of artefacts that are associated with spiral-CT imaging. A simple description of spiral-CT reconstruction is used to explain how these artefacts depend on the pitch and subject morphology, and shows when these artefacts are likely to impair the diagnostic value of the acquired images. We scanned a cone and rod phantom with pitch 2, and used the acquired images to demonstrate how spiral data acquisition and interpolation leads to artefacts in the reconstructed images. We then demonstrated the effects of various pitches in scans of a human cadaver, whereas the slice thickness was kept constant. Some patient studies are presented in order to show the possible clinical consequences. Spiral acquisition may cause geometric distortions and apparent inhomogeneity of homogeneous structures. We were able to link these artefacts to the way in the acquisitions were done, and the reconstructions were performed. We have shown how these artefacts can be anticipated in clinical studies. When areas of low contrast, surrounded by hypo- or hyperdense structures, are scanned with a large pitch and viewed with a narrow window, spiral artefacts may influence the diagnostic quality of the images. These effects should be considered when choosing the pitch.

SCTA of the renal arteries was best performed with a scan delay of TTEST + 20 s. However, analysis of our data showed that similar results could be expected with a delay of 44 s.

No abstract