Four-dimensional speckle tracking for strain imaging in mice

The prevalence of medical imaging techniques in the diagnostic protocol of several structural heart diseases (SHDs) has led to the increased usage of structural indices, such as the global longitudinal strain (GLS) in determining cardiac dysfunction. Such structural indices obtained through the analysis of cardiac motion have been found to be important in the early-stage assessment of several SHDs. Organ-level measures such as GLS do not capture the complexity of cardiac motion, which is characterized by significantly heterogeneous variations in contractile patterns. Additionally, GLS is often estimated using brightness-mode (B-mode) echocardiography, which is challenged by the rapid motion of the heart in small animal models such as mice. In this study, we hypothesize that four-dimensional (4D) US imaging through the implementation of a high-frequency transducer will provide a complete spatiotemporal description of 4D cardiac motion. The left ventricle (LV) of a murine heart was segmented as point clouds at various time points across the cardiac cycle between end-diastole (ED) and end-systole (ES). The iterative closest point algorithm was used to perform point-set registration to estimate displacements and strains at ES relative to ED. The proposed methodology was first validated against benchmark, speckle-tracking-derived displacements in synthetic US images. Our findings evidenced qualitative similarity in the displacement maps derived at various planes of the phantom using the proposed method and speckle tracking. Our proposed methodology provides a rigorous protocol to quantify complete spatiotemporal cardiac motion using readily available imaging modalities, aiding in the much-needed adoption of regional strain markers in the clinic.