Reproducibility of noncalcified coronary artery plaque burden quantification from coronary CT angiography across different image analysis platforms

OBJECTIVE. The objective of our study was to evaluate the reproducibility of noncalcified coronary artery plaque burden quantification from coronary CT angiography (CTA) across different commercial analysis platforms. MATERIALS AND METHODS. For this study, 47 patients (36 men, 11 women; mean age ± SD, 62 ± 13 years) with noncalcified plaques on coronary CTA were included. Automated quantification of non-calcified coronary artery plaque volume was performed on identical datasets using three commercially available image analysis software platforms (software platforms 1-3). Identical tissue attenuation ranges between 0 and 50 HU for low-attenuation plaques and 50-130 HU for medium-attenuation plaques were consistently applied. Log volume data were compared with the Pearson correlation coefficient and Bland-Altman analysis. RESULTS. Differences in plaque volume measurements on intraplatform repeat measurements were statistically insignificant (p = 0.923). At the low-attenuation threshold, software platform 3 had significantly higher log volumes (p < 0.001) than both software platforms 1 and 2 and software platform 1 had significantly higher log volumes than software platform 2 (p < 0.001). The results at the medium-attenuation level were identical except that the log volumes for software platforms 1 and 2 were not significantly different (p > 0.05) in the left anterior descending artery and left circumflex artery. The Pearson correlation coefficient was found to be 0.677 (p < 0.001; 95% CI, 0.608-0.735) between software platforms 1 and 2, 0.672 (p < 0.001; 95% CI, 0.603-0.732) between software platforms 1 and 3, and 0.550 (p < 0.001; 95% CI, 0.463-0.627) between software platforms 2 and 3. CONCLUSION. Currently available noncalcified plaque quantification software provides good intraplatform reproducibility but poor interplatform reproducibility. Serial or comparative assessments require evaluation using the same software. Industry standards should be developed to enable reproducible assessments across manufacturers.