Studies of CT dose and image quality using clinical and phantom images

X-ray computed tomography (CT) has shown remarkable technological developments and advances in performance in the past two decades, and is one of the prime established radiologic modalities today. CT examination is associated with high quality diagnostic images, but also with high radiation doses. This thesis examined the radiation doses from CT machines, the relationship between CT image quality and radiation dose and between image quality and noise using clinical CT images and phantom acquisitions. The first study used non-randomised clinical image data collected from three hospitals on single, 16-and 64-slice CT scanners. A total of 240 patients who underwent abdomen and brain scans were used for image quality evaluation and dose assessment. The image qualities were evaluated by four human observers (two radiologists and two technologists). The two radiologists and the two technologists each had 10 years' experience, and used a one-to-five visual grading scale. The radiation dose index (CTDIv) was either recorded from the image display or calculated by the ImPACT software using the recorded scan parameters. The averaged CTDIv were 32.2, 46.0 and 69.8 mGy for the brain and 9.5, 14.5 and 16.9 mGy for the abdomen from single-, 16- and 64-slice scanners respectively. The averaged image quality scores were 2.71, 3.32 and 4.35 for single, 16, and 64-slice scanners. Single-slice CT scanners delivered the least radiation dose to patients in comparison with the 16- and 64-slice CT scanners, and the image quality was adequate for diagnosis. Therefore, both 16- and 64-slice CT scanners delivered image quality that was more than adequate, as well as higher than necessary radiation doses to patients. The patient dose from the 16- and 64-slice scanners may be reduced by lowering their image quality to the level of a single-slice CT scanner.