Abstract 17344: Reproducibility of Transluminal Attenuation Gradient Measurement Using Semi-automated Coronary Computed Tomography Quantification Software
Background: Transluminal attenuation gradient (TAG) provided by coronary computed tomography angiography (CCTA) is a method to predict functionally significant coronary artery stenosis and previous studies demonstrated the feasibility of assessing TAG. However, limited research has been performed to evaluate the reproducibility of TAG measurement between readers. We investigated inter-observer variability of TAG using a semi-automated CT quantification software.
Method: Forty six patients who underwent clinically indicated CCTA were evaluated. TAG values were measured using semi-automated CT quantitatification software (QAngio) and defined as the linear regression coefficient between luminal attenuation and length in coronary vessels. Two experienced observers independently measured each coronary artery. Inter-observer variability by different readers was examined for the TAG measurements.
Result: Sixty four vessels (LAD: 30, LCX: 18, RCA: 16) were assessed. Vessels without stenosis, total occlusion, or stent were excluded. Among the first and second observers, there was no significant difference between overall TAG (-1.401±1.1 vs. -1.396±1.1 HU/mm, p=0.98). Furthermore, we divided the analysis into each of the 3 major coronary arteries, which also showed no significant difference (-1.413±0.9 vs. -1.408 ± 0.9 HU/mm , p=0.98 for LAD, -2.197±1.2 vs. -2.183±1.2 HU/mm, p=0.97 for LCX, -0.482±0.6 vs. -0.489±0.6 HU/mm, p=0.98 for RCA). Bland-Altman analysis showed that the TAG measurements were highly reproducible (r=0.99, p=0.82 for overall, r=0.99, p=0.86 for LAD, r=0.98, p=0.80 for LCX, r=0.97, p=0.87 for RCA).
Conclusion: The present study shows excellent correlation of TAG values for both coronary vessels and for each major coronary artery vessels between two readers. The semi-automated CT quantification software provides a reproducible approach to assess TAG.
Author Disclosures: S. Matsumoto: None. R. Nakanishi: None. A. Alani: None. M.A. Fahmy: None. J. Abraham: None. D. Li: None. C. Dailing: None. F. Flores: None. P. Kitslaar: None. A. Broersen: None. M.J. Budoff: None.
- © 2014 by American Heart Association, Inc.