SU-F-207-08: CT Protocol Management Using Quantitative Clinical Image Quality Measurement




To establish a CT protocol management program based on quantitative measurement of clinical Image Quality.


After reviewing images of routine adult brains and abdomens for 12 system scanners, we established a target Image Noise range. The scanners are of different vendors, models and configurations. For instance, only 4 have iterative reconstruction capability. For this reason it is problematic to rely on dose alone as a standardization metric.We then instituted quarterly review of clinical images. A set of 10 images is randomly selected, and reviewed to determine if the target noise range is being achieved. This is done for both routine adult brains and abdomens. Outliers are studied to determine if further action is needed, e.g. protocol revision. The performance of the sample as a whole is also examined.


The program was implemented spring of 2014. To date, it has successfully identified situations where Image Noise systematically exceeded the target range. Interventions were made in each one of these cases, successfully adjusting the Image noise to within desired limits. In addition, the program is used as a Measure of Effectiveness for proposed protocol revisions, producing an objective performance measure.


Measuring and monitoring clinical image noise is a simple, effective, quantitative tool for managing CT protocols.