Reproducibility of T1 mapping measurements of interventricular septum in short term follow up on 3T MRI scanner

Abstract

Background / Objective T1 mapping is widely used for tissue characterisation of the myocardium in cardiac MRI scans. Additionally, it can be a quantitative criterion in follow-up scans to assess interval change. High reproducibility of T1 mapping is crucial to ensure its usefulness in clinical practice as a follow-up assessment. However, the T1 mapping values depend on various factors (MRI field strength, scan sequence, heart rate, etc.) and significant artifacts may occur. The aim of this study was to assess the reproducibility of myocardial regional T1 mapping measurements in the interventricular septum using identical scan parameters and slice positions over a short time period, which on the 3T scanner is quite sparse elucidated. Methods Study cohort consisted of 51 patients who underwent cardiac MRI on 3T scanner. Native T1 mapping images were acquired in short-axis plane at three levels (basal, middle, and apical locations) in two time points: after scout sequences acquisition and at the end of the study or before intravenous contrast injection. The identical planes and saturation region for these measurements were applied. T1 mapping measurements were conducted by placing identical ROI’s of 1cm2 in the identical interventricular septum locations. Results The mean time interval between the two T1 mapping acquisitions was 19.4±5.6 minutes. No statistically significant difference between two T1 mapping values were observed. Reproducibility for interventricular septum was excellent for all slice position, with the highest one in the midventricular slice (intraclass correlation [ICC] 0.95), followed by basal slice (ICC 0.93)) and apical slice (ICC 0.92). F-test (0.7) did not find a statistically significant difference in the ICCs for the midventricular and apical slices. Conclusions T1 mapping values in a short time follow-up demonstrated excellent reproducibility in all locations of interventricular septum, although the highest one was observed in the midventricular slice and the lowest in the apical slice. Our findings are in line with results of Dekkers et. al. (J. MAGN. RESON. IMAGING 2019), which also found best reproducibility in the midventricular slice of septum, although our study demonstrated better intraclass correlation and lower difference of the reproducibility to the basal and apical septal slices.