The Effect of Smoothing and Drift Compensation on Fetal Strain

Objective

The aim of this study was to assess the effect of user-regulated image-processing settings (spatial smoothing, temporal smoothing and drift compensation) on fetal left ventricular strain.

Methods

Left ventricular average longitudinal strain was acquired from the four-chamber view of the fetal heart from 34 fetuses, with 30 fetuses presenting adequate quality. A total of 18 different settings for spatial smoothing, temporal smoothing and drift compensation were examined. At each setting the average strain for the 30 fetuses was calculated, whereby one could examine whether there was an average difference in fetal strain at the different settings. Furthermore, the difference between the highest and lowest strain values across the 18 settings was assessed for each fetus (min-max difference). The average min-max difference was then calculated across the 30 fetuses to calculate the mean discrepancy in fetal strain due to smoothing settings.

Results

The average effect of the smoothing settings as well as drift compensation by them was small. However, when examining the discrepancy induced by the different settings together, they induced average proportional differences of approximately 18% for the endocardial and epicardial layers and 15% for the mid-wall layer.

Conclusion

This study shows that while the average effect of different smoothing settings and drift compensation was small, they induced significant discrepancy in strain values on the individual level. We recommend that examiners be consistent with regard to smoothing and drift compensation settings.