Left ventricular strain analysis-the importance of being expert

Abstract Funding Acknowledgements Type of funding sources: None. OnBehalf - Background Myocardial strain imaging using speckle-tracking echocardiography is widely used in both research and clinics. Left ventricular global longitudinal strain (GLS) has proven to be a reproducible and useful tool in clinical practice. However, data about the variability in global and segmental strain among novice are limited. Purpose The aim of this study was to evaluate how the variability of strain measurements depends on the experience of the physician performing the analysis. Furthermore, we tried to assess how image quality and presence of pathology influence the diagnostic performance of the analysis. Methods Twenty novice (no or limited experiences with strain analysis) and 5 experts analysed offline 3 cases: a healthy adult and a patient with ischemic cardiomyopathy, both with high image quality, and a healthy adult with suboptimal image quality (insufficient tracking in two segments). Frame rates were 65, 51, and 70 fps, respectively. Left ventricular GLS and segmental longitudinal peak systolic strain were quantified using the automated function imaging protocol from vendor-specific offline analysis software. The absolute difference in GLS and segmental strain between each novice readers and experts was calculated. Mean strain measurements from the experts were used as a comparator. Results Absolute values of GLS ranged from -13.6% to -20.1% in the novice group, while GLS ranged from -15.6% to -18.8% in the expert group. The absolute difference in GLS was significantly higher in the novice group compared to the expert group (P < 0.001). Absolute differences in GLS varied significantly among cases, with low variability in healthy adult with high image quality (Panel A) and with significantly higher variability between novices and experts in case with suboptimal image quality and ischemic cardiomyopathy (Panel B-C). The absolute differences in segmental longitudinal peak systolic strain were up to 11.9% strain unit in novice group and up to 6,8% strain unit in experts (P < 0.001 between groups). In novice group, highest absolute differences in segmental strain were observed in the region with suboptimal tracking and in infarct region in patient with ischemic cardiomyopathy (Panel D-E). Conclusion Left ventricular strain analysis by a reader with no or limited experiences severely affects the diagnostic potential of this method. Significant variability in strain measurements should be considered especially in cases with suboptimal image quality and in cases with regional left ventricular pathology. Abstract Figure.