2D Strain: The Choice of the Width of the Region of Interest Has in Itself a Significant Influence on Measured Strain

Introduction: To implement speckle tracking-derived global longitudinal and global circumferential strain (GSL, GSC) in daily clinical practice, it is crucial to identify factors which are influencing the reliability of the measurements. To obtain GSL and GSC, a region of interest (ROI) has to be created. The operator selects the width to include the whole LV myocardium while excluding the pericardium. The purpose of this study is to test the hypothesis that the width of the ROI has, in itself, a significant influence on the measured values.

Methods: In 20 healthy subjects, two cineloops were analyzed, an apical four-chamber view (4CH) and a parasternal short axis view at the level of the mitral valve (SAX). A ROI was created by the operator. In a first step, the width of the ROI was set automatically by the software (medium ROI). Two additional measurements were obtained from each cineloop by making the width of the ROI, one step narrower (narrow ROI) and another one step wider (wide ROI) than the medium ROI.

Results: The mean values for the narrow, medium, and wide ROI were for the GSL –21.9%± 1.9, –20.1%± 1.7, and –19.2%± 1.6 and for the GSC –21.9%± 2.6, –19.8%± 2.5, and –18.1%± 2.3. The mean differences between the measurements of the narrow and the medium ROI and between the medium and wide ROI were for the GSL –1.8% (–2.1 to –1.5) and –0.9% (–1.1 to –0.7) and for the GSC –2.2% (–2.5 to –1.9) and –1.7% (–2.0 to –1.4) (all p < 0.000).

Conclusion: GSL and GSC values vary significantly when the width of the ROI is changed by the operator. Accurate ROI width definition is essential. High quality images with clearly identifiable endo- and epicardial borders are critical to allow accurate measurement.