Background: To determine the normal range of left ventricular (LV) 2- dimensional (2-D) strain and vendor-specific differences, a multicenter prospective 2-D strain study endorsed by the Japanese Society of Echocardiography was conducted.
Methods and results: 2-D speckle tracking analysis was performed on 817 healthy subjects (age range, 0-88 years); the images included 3 LV short axis and 3 apical views using an ultrasound system from 1 of the 3 different vendors (V(1), n=333; V(2), n=330; V(3), n=337). With the 2-D speckle tracking software from each vendor, radial, circumferential and longitudinal strain were measured using an 18-segment model. Inter-vendor variability was also assessed in a subset of subjects. The feasibility for 2-D strain measurements was different among the 3 vendors (V(1), 83%; V(2), 70%; V(3), 88%, P<0.01). The global radial (V(1), 54.6±12.6%; V(2), 36.3±8.2%; V(3), 51.4±8.0%), circumferential (V(1), -22.8±2.9%; V(2), -22.2±3.2%; V(3), -30.5±3.8%), and longitudinal (V(1), -21.3±2.1%; V(2), -18.9±2.5%; V(3), -19.9±2.4%) strain measurements were significantly different for each of the vendors. Segmental strain was also different between the 3 vendors. On inter-vendor analysis, vendor agreement ranged from mild to moderate.
Conclusions: Reference values are provided for normal 2-D strain for 3 different ultrasound vendors. Due to a low inter-vendor agreement, 2-D strain data are not interchangeable when conducting a longitudinal follow-up or a cross-sectional assessment of LV function.