Abstract
Funding Acknowledgements
Type of funding sources: None.
Background
Optimal 3D speckle tracking echocardiography (STE) requires volume stitching, a technique prone to errors, such as patient movements, respiration and arrhythmias. With different resolution at different depths (Figure 1A), such errors may cause stitching artefacts that affect STE derived deformation measurements at different levels of the left ventricle.
Methods
In this cross-sectional study, 62 patients with variable degree of aortic stenosis participated. 3D images were recorded using 3 different methods (2-, 4- and 6-beat volume stitching at 22 ± 2, 29 ± 4 and 36 ± 6 volumes per second, respectively). Strain, rotation and torsion were analyzed at 3 different levels (basal, mid and apical) (Figure 1B). Results were compared using one-way analysis of variance between methods. Inter-segmental variation was estimated using the coefficient of variation (CV) of the segments involved for each level. Results were interpreted as more accurate if low values of inter-segmental variation between segments in a level was present.
Results
Neither longitudinal strain (LS) nor circumferential strain (CS) showed any difference between methods in the basal level (Table 1). LS was higher in mid and apical level for 2-beat, while CS was higher only in apical level for 2-beat.
Inter-segmental CV showed no significant difference between methods for LS and CS. Between levels, both had higher CV in basal than apical level, but CS had much higher values, indicating lower accuracy in basal level.
Rotation measurements showed negative values in basal level and positive in apical level. None of the methods differed significantly. Torsion showed higher values for 2 beat images only.
Conclusion
3D STE from both 4- and 6 beat images have insignificant differences for strain, as well as low inter-segmental variability. LS is accurate in all levels while CS has high accuracy in apical, and poor in basal level. Both rotation and torsion differ insignificantly when using 4- and 6-beat images.
Table 1 Level Mean ± SD ANOVA P CV 2-beat 4-beat 6-beat 2-beat 4-beat 6-beat Longitudinal strain (%) Basal -17.82 ± 4.83 -16.81 ± 4.48 -17.05 ± 3.56 0.402 34.5 34.9 31.5 Mid -16.34 ± 4.30 -14.91 ± 3.62 -14.46 ± 3.53 0.019 31.4 29.9 29.9 Apical -16.80 ± 4.33 -15.06 ± 3.12 -15.09 ± 3.02 0.008 30.6 28.5 24.9 Circumferential strain (%) Basal -12.93 ± 4.42 -14.16 ± 4.21 -13.05 ± 3.24 0.168 106.7 71.7 74.4 Mid -17.59 ± 3.86 -16.73 ± 3.56 -16.98 ± 3.51 0.406 43.0 36.9 30.3 Apical -22.14 ± 6.20 -19.04 ± 4.81 -20.32 ± 5.40 0.008 26.8 29.4 23.4 Abstract Figure 1