CT Users Group

Nicola Bate

Royal Infirmary of Edinburgh

Abstract

PURPOSE: To compare image quality in a number of CT examinations for radiology departments across NHS Lothian and NHS Fife, and establish whether there is any correlation with dose, identifying potential for dose optimisation.

MATERIALS/METHODS: Dose audits were carried out across the region, whereby DLPs were collected for patients of standard weight 50-90kg. This gave an average DLP for each scanner for three examinations (Head, Chest-liver, Abdo-pelvis). Noise levels in clinical images were also audited by accessing patient images. Noise was measured as the standard deviation of CT number in the descending aorta (chest/abdo) or ventricles (head). Both DLPs and measured noise were then compared between seven scanners.

RESULTS: We had expected that DLP values vary from scanner to scanner, however we had hoped that these differences could be justified by equal image quality. The study found that radiologists were accepting a significant range of image quality across the region. Noise levels varying significantly can be attributed to a combination of factors, including imaging parameters, directly related to dose (e.g. tube current) and those not directly related to dose (e.g. kernel, slice thickness).

CONCLUSION: At centres where higher DLPs are leading to below average noise levels the image quality data provide evidence to challenge whether these higher dose protocols are clinically justified. Changes could be made to protocols to ensure images quality is consistent across the region and doses ALARA. This approach offers a further step in dose-image quality optimisation beyod simple comparison against existing DRLs.

Click the link for a pdf of the presentation:

Back to meeting timetable