| Ultrasound strain imaging has been proposed to quantitatively assess myocardial contractility. Cross-correlation based 2D speckle tracking (ST) and auto-correlation based Doppler tissue imaging (DTI) are competitive ultrasound techniques for this application. Unlike ST, DTI is limited by angle dependency and suffers from low signal-to-noise ratio because a high pulse repetition frequency (PRF) is used to reduce decorrelation due to lateral or elevational motion. However, axial normal strain estimated using these methods have not been quantitatively compared on a well-controlled animal model. To address this, we tested both methods on two rabbit hearts: one control experiment and one ischemic heart created by ligating the left descending anterior (LAD) artery. Both rabbit hearts were retroperfused through the aorta with oxygenated Krebs-Henseleit buffer and paced through the apex at 3Hz. A linear probe (Sonix RP) were used to acquire radio frequency (RF) data at a PRF appropriate for DTI (740Hz) with three kinds of transmit pulses P1, P4 and P12 (1, 4 and 12 cycles, respectively). Acquisition of RF data, left ventricular pressure, and electrocardiograms were synchronized to the pacing signal. 2D ST estimated strain had a resolution of 1.2 mm axially and 2.5 mm laterally for transmit pulse P1. Color Doppler imaging estimated strain only along the beam direction (axial direction) were derived from the velocity filtered to achieve the same spatial resolution as axial displacement estimated using ST with transmit pulse P1. Compared to 2D ST, strain using DTI can achieve high temporal resolution but low strain signal to noise ratio. The spatial resolution and detection accuracy also depend on transmit pulse length (bandwidth). Both axial normal strain using ST and M-mode DTI can detect wall thinning (abnormal motion) during systole due to LAD ligation on the second heart. Unlike ST, DTI derived strain imaging underestimated abnormal motion due to incorrect lateral registration during accumulation. |