Intra‐ and interreader reproducibility of magnetic resonance imaging for quantifying the lipid‐rich necrotic core is improved with gadolinium contrast enhancement

Abstract

To test the hypothesis that intra- and interreader reproducibility for measuring the lipid-rich necrotic core (LR-NC) size is significantly improved with gadolinium (Gd) contrast-enhanced magnetic resonance imaging (CEMRI) compared to non-CEMRI. Thirty-seven individuals with >50% carotid artery stenosis underwent carotid MRI at 1.5T (pre- and postcontrast T1-weighted (T1W), T2-weighted (T2W), proton density-weighted (PDW), and three-dimensional time-of-flight (TOF) sequences). Two independent readers measured the mean area of the LR-NC from the precontrast images only, followed by a second measurement using the additional postcontrast images. One reader repeated the measurements after an interval of five months. Intra- and interreader reproducibility was analyzed by means of the intraclass correlation coefficient (ICC), coefficient of variation (CV), and standard deviation (SD). The CV decreased from 33.7% to 8.8% for intrareader measurements of the LR-NC, and from 33.5% to 17.6% for interreader measurements. The SD was significantly smaller with CEMRI than with non-CEMRI (P = 0.003 and P = 0.006, respectively). The ICC increased from 0.94 to 0.99 and from 0.85 to 0.93 for the intra- and interreader measurements, respectively. Reader reproducibility for in vivo MRI quantification of LR-NC size is significantly improved by the addition of Gd contrast in individuals with >50% carotid stenosis.