Quantification of intracranial internal carotid artery calcification on brain unenhanced CT: evaluation of its feasibility and assessment of the reliability of visual grading scales

Abstract

Previous studies have evaluated intracranial internal carotid artery calcifications (ICACs) qualitatively using different visual grading scales, which could lead to inconsistent results. The purpose of this study was to evaluate the feasibility of ICAC volume measurement and to correlate the volume with visual grading scales. We included 49 patients (>50 years) who underwent unenhanced cranial CT. Two observers evaluated four visual grading scales and measured ICAC volumes semi-automatically, and interobserver agreements were assessed. Differences in ICAC volume between visual grades of each scale were tested. The relationship between the visual grading and volume was assessed. Interobserver agreements ranged from 0.841 to 0.901 for visual grading and 0.997 for volume measurement. Mean volumes were not significantly different between the visual grades (P > 0.05) except when comparing grade 4 with the other grades. The grades of each visual grading correlated well with ICAC volumes (Spearman's ρ = 0.849-0.881, P < 0.001). The relationship between the visual grades and volume was described by a quadratic model (R (2), 0.31-0.50, P < 0.01). ICAC volume measurement is feasible and reproducible, whereas visual grades poorly reflect the actual volume; therefore, volume measurement may be warranted for future research.