Inter- and intra-rater reproducibility of semiautomatic determination of volume parameters in cardiac magnetic resonance imaging

Abstract

Purpose: The purpose of this study was to evaluate inter- and intra-rater reproducibility in volume assessment using cardiac magnetic resonance imaging (CMRI).

Methods: Twenty-five healthy volunteers and 106 patients were included into this retrospective study and received CMRI. The patients were divided in three groups (group I, 80 patients with arrhythmia; group II, 20 patients with cardiomyopathy; group III, 6 patients after correction of septum defects). Therefore, the images were semiautomatically segmented by an experienced and an unexperienced radiologists. The analysis of end-diastolic volume (EDV), end-systolic volume (ESV) and stroke volume (SV) as well as ejection fraction (EF) and myocardial mass (MM) were performed twice by an experienced and an unexperienced radiologists. The intra-class correlation coefficients (ICC) were determined for the evaluation of inter- and intra-rater variance.

Results: The intra-rater reproducibility for determination of EF, ESV, EDV and MM was excellent with ICCs ranging from 0.88 to 0.99 (all p<0.001). The inter-observer reproducibility for these parameters was also excellent with ICCs ranging from 0.91 to 0.98 (all p<0.001). The assessment of the SV showed an excellent intra-rater agreement with ICCs of 0.96 and 0.92 (both p<0.001), but only a moderate ICC for the inter-rater reproducibility (0.54, p<0.001).

Conclusions: Our study shows that assessment of cardiac volumes can be performed on CMRIs with an excellent reproducibility by both experienced and unexperienced investigators.