Strain rate imaging: an in vitro "validation" study using a physiologic balloon model mimicking the left ventricle.

ABSTRACT

BACKGROUND: Strain rate imaging (SRI) can be implemented from digital ultrasound loops of tissue Doppler imaging (TDI) data and is performed as an autocorrelation solution of the distance between intramyocardial targets. As such, it should have better resolution along longer distances of wall segments that are imaged at the length of individual ultrasound scan lines. METHODS: We used a new left ventricular double-balloon phantom with a tissue-mimicking gel between the walls. Mounted in a water bath and connected to a pulsatile flow pump at four-stroke volume (30-50 ml/beat), the high frame rate, digital, multiple two-dimensional/tissue/TDI loops of balloon wall motion were recorded using a GE VingMed system FiVe (3.5 MHz phased array transducer), with the model scanned longitudinally from the apex. The strain rate (SR) values were measured at the apex and the lateral wall using an offline measurement program, and mean SR values for every 100 msec were calculated by averaging three determinations at each point. The excursions of the apex and lateral wall also were measured directly by high speed digital video imaging, and consecutive velocity profiles were calculated every 100 msec. A total of 40 data points for four-stroke volumes were analyzed. RESULTS: While our balloon model had enough gel targets between the walls to produce a good mimic of myocardial speckle with walls that thickened and thinned, samples immediately across the apex and apex SR values (Hz) varied substantially. In contrast, systematic signals could be obtained from lines imaged >15 degrees from the true apex and crossing a longer length of myocardium. At the lateral wall, there was a close correlation between the video velocities and SR values, as well as a close overlap of the phasic patterns. CONCLUSIONS: SRI produces more reliable data from wall segments parallel to scan lines.