A new structured, dose-aware 5-star image quality rating system for CT scans has shown robust reproducibility across six European hospitals, according to a study published August 6 in the European Journal of Radiology.
A multinational group coordinated through the International Atomic Energy Agency (IAEA) evaluated the system on CT exams across six hospitals in five European countries, with findings suggesting it can standardize dose-aware CT quality assessment and identify optimization opportunities, noted lead author Madan Rehani, MD, of Harvard Medical School in Boston, and colleagues.
“We found that radiologists rate CT image quality consistently across different institutions and countries, which shows the 5-star system is genuinely reproducible in real practice, not just in theory,” Rehani told AuntMinnie.
The absence of an internationally harmonized method for rating CT image quality in routine practice carries measurable consequences, including unwarranted radiation exposure to patients from repeat imaging and suboptimal protocol design, Rehani and colleagues explained. The group previously formalized the 5-star framework to address this gap. Notably, the 5-star system explicitly inverts the traditional assumption that lowest image noise equals best quality, emphasizing instead that unjustifiably low noise (typically implying higher radiation dose) should be rated down, not up.
“The crispest image is often the one that uses more radiation than the diagnosis actually needs,” Rehani said. “Our 5-star system scores that image a 2, not a 5, so radiologists can flag when quality is too good for the dose it cost, and reward images that are good enough at a lower dose instead.”
To test the rating framework’s feasibility and reproducibility, the researchers conducted a retrospective analysis of 2,737 adult CT exams (1,247 chest and 1,490 abdomen) across 13 scanners at six hospitals in five European countries. Three radiologists per institution independently scored each exam after a structured orientation, with inter-reader agreement, star-rating concordance, and dose metrics (volume CT dose index [CTDIvol] and dose length product [DLP]) evaluated per hospital and clinical indication.
The rating framework demonstrated robust reproducibility, with three-reader concordance reaching up to 93% in highly standardized centers and complete discordance remaining exceptionally rare (1.6% cases overall). The researchers noted a compelling inverse correlation between CTDIvol and star ratings, especially among larger patients, with higher doses consistently corresponding to lower star ratings.
In addition, noise-tolerant indications (CT for pulmonary nodules and kidney stones) showed higher-than-expected radiation exposure, highlighting an opportunity for optimization, and certain centers ran higher-dose protocols that skewed heavily toward 2-star ratings, the researchers reported.
“The 5-star CT [image quality] rating system is reproducible and supports dose-aware CT optimization,” the group wrote.
The next step is for departments to actually put the 5-star system into routine use, starting with readers flagging individual exams that look unnecessarily low noise for the dose they cost, Rehani noted. From there, departments could compare their own ratings and doses across indications to see where local protocols need adjusting.
“We would also like to see it tested prospectively, ideally alongside newer reconstruction and AI based tools,” Rehani said.
The full study is available here.