P5288Slice position vulnerability in the basal and apical parts for right ventricular circumferential strain measurement with feature tracking cardiac magnetic resonance
Abstract Background Strain is a more sensitive and precise parameter than ejection fraction (EF) for detection and characterization of subclinical left ventricular (LV) dysfunction and remodeling. Similar relationship is expected for right ventricle (RV); however RV functional parameters are less validated. Feature tracking strain analysis based on standard cardiac magnetic resonance (CMR) cine imaging is available for both ventricles. We experience a large slice-to-slice variation for RV global circumferential strain (GCS), possibly making the parameter vulnerable to minute position changes. Purpose To evaluate slice-to-slice differences in RV GCS for identification of the least variation region in a patient group without regional RV disease, in order to achieve a robust method for measurement. Hypothesis The slice-to-slice difference in peak GCS is lower in the mid-ventricular part of the RV than in the basal and apical parts. Methods 50 patients 6–72 months after pulmonary embolism without other major cardiopulmonary disease were included; mean age 60 years (range: 18–75 years); 68% men. Standard 2D cine CMR was obtained in longitudinal planes and in 10–12 consecutive 10 mm short axis planes for complete coverage of the RV. RV free wall and the inner contour of the septum were manually segmented on every end-diastolic and end-systolic slice from the pulmonary valve to the apex for feature tracking strain analysis. Peak RV GCS for every short axis slice and GCS difference (absolute percentage points) between adjacent slices were calculated. RV EF and peak RV GLS from the 4-chamber image were measured for correlation to RV GCS. Wilcoxon signed rank test and Pearson correlation were performed. Confidence intervals of means are based on 1000 bootstrap samples. Results RV EF was 46.6% (95% CI: 44.3; 48.8), RV peak GLS was −17.6% (95% CI: −18.6; −16.6). RV mid-ventricular GCS was −10.9% (95% CI: −12.0; −9.9). RV peak GCS slice-to-slice difference was 6.8 absolute percentage points (95% CI: 6.0; 7.6) in the basal part, 2.7 (95% CI: 2.4; 3.0) in the mid-ventricular part and 4.6 (95% CI: 3.9; 5.3) apically. Difference was significantly lower in mid-ventricular (p<0.001) compared to both basal and apical. RV EF correlated to RV peak GLS (r: −0.397, p=0.004) and mid-ventricular peak GCS (r: −0.356, p=0.01) but not to basal or apical peak GCS. RV peak GLS correlated to basal and mid-ventricular peak GCS (r: 0.313, p=0.03 and r: 0.301, p=0.03 respectively) but not to apical peak GCS. Figure 1 shows slice-to-slice difference (expressed in absolute percentage points) in right ventricular peak GCS. Conclusion Slice-to-slice difference in RV peak GCS was significantly lower in the mid-ventricular region. Large differences in the basal and apical parts indicate that measurements largely depend on slice positioning.