Abstract 12968: Disparity of Strain Measurements Using Single Vendor Analysis Programs
Background: Serial strain measurements for detection of early, subclinical disease are being advocated by some guidelines, which advise modification of treatment based on these results. Inconsistency of strain data between different vendors is problematic and acknowledged. We sought to determine if this disparity in strain measurement also occurs using different analysis programs from the same vendor.
Methods: Offline analysis was performed to assess 2D longitudinal strain on 20 consecutive subjects. Apical views (4-chamber [4ch], 2-chamber [3ch], and 3-chamber [3ch]) were used to obtain regional and mean global strain (GLS). Analysis was performed using commercial software (EchoPAC BT12; GE Healthcare, Horten, Norway) with semi-automated tissue tracking. Strain measurements on the same images were then repeated with a newer version of the software (EchoPAC, BT13) featuring automatic tissue tracking and using vendor-neutral software (EchoInsight, Epsilon Imaging, MI). Strain data was then compared using paired T-tests and analysis of variance.
Results: EchoPAC BT12 and BT13 strain values were well correlated (mean GLS r=0.929, 4ch r=0.94, 2ch 0.879, 3ch r=0.912) with p<0.001 for all. However, BT12 strain results were consistently larger (in absolute value) with a mean GLS of -16.7±4.7% compared with a BT13 mean GLS of -15.6±4.6% (p=0.014). Strain results using vendor-neutral software demonstrated good correlation with both vendor specific programs (B12 mean GLS r=0.924, B13 mean GLS r=0.921; p<0.001 for both) and were not significantly different from vendor specific strain values (p=0.75). [Table]
Conclusions: Disparity in strain measurements performed using different analysis programs from the same vendor is important to consider when assessing serial strain measurements. This may have a significant impact on initiation or withdrawal of therapy. Vendor-neutral strain programs may overcome these limitations in sequential measurement.
Author Disclosures: C. Jellis: None. Z. Popovic: None. C. May: None. J. Thomas: None.
- © 2014 by American Heart Association, Inc.