CT "5-Star" Image Quality Scoring May Help Reduce Radiation Dose

A multinational study involving the International Atomic Energy Agency and Massachusetts General Hospital in the United States shows that a standardized "5-star" scoring system that considers both radiation dose and CT image quality can produce relatively consistent evaluations across different radiologists and help identify opportunities to reduce patient radiation exposure.

This retrospective study included 2,737 adult CT examinations performed at 6 hospitals across 5 European countries, including 1,247 chest CTs and 1,490 abdominal CTs, involving 13 CT scanners. At each hospital, 3 radiologists independently evaluated the examinations using the "5-star" framework.

Unlike traditional image quality assessments that may lean toward "the lower the noise, the better," this system also focuses on whether the radiation dose used to acquire the images is reasonable. In other words, if an image has very low noise but is based on unnecessarily higher radiation exposure, its score may actually be reduced.

The study showed that inter-reader agreement was generally strong. At centers with more standardized workflows, the agreement rate among the 3 readers reached up to 93%; overall, completely discordant ratings accounted for 1.6% of cases.

The researchers also found an inverse relationship between CTDIvol, a metric reflecting scanner radiation output, and star ratings, which was more pronounced in larger patients. Dose levels also varied significantly across different scanners and participating institutions.

The analysis also indicated that some CT examinations could have their scan protocols further adjusted based on clinical indications. For example, pulmonary nodule and kidney stone examinations can typically tolerate higher image noise while maintaining diagnostic value, yet some cases still used relatively higher radiation doses. Some centers also more frequently used high-dose protocols, resulting in a higher proportion of 2-star ratings.

The study authors believe this approach can serve as a complement to diagnostic reference levels based on clinical indications, helping healthcare institutions compare CT practices across different scanners and sites. Potential optimization areas include radiation output, the number of image acquisition sequences, and image reconstruction methods.

The study, titled "Dose-Aware 5-Star Image Quality Scoring in CT: Results from the IAEA-MGH Study," has been published in the European Journal of Radiology.