High frame rate speckle tracking algorithm towards a real-time implementation

High frame rate speckle tracking echocardiography (HFR STE) assesses clinical biomarkers in a user-friendly manner. Real-time feedback of the estimated motion may improve acquisition quality, speed up examinations, and reduce patient discomfort. However, its implementation is challenging due to the computational load requirements. The aim of this study was to improve the computational efficiency of a previously presented HFR STE algorithm to make feasible its real-time implementation on the ULA-OP 256 research scanner. The axial and lateral motion estimators were adapted to IQ demodulated data, and modified to decrease the computational load. To validate the implementation, apical 4-chamber views were recorded from healthy volunteers and the motion results obtained in real-time were compared offline with those produced by the original approach. In both cases, the same myocardial contour was tracked to extract the velocity, strain, and strain rate curves. The extracted curves had a physiological pattern and together with the clinical biomarkers were not significantly different from those obtained with the original approach.