In clinical ultrasound, conventional 2-D strain imaging often struggles to accurately quantify the heart’s three orthogonal normal strain components. This limitation arises from the need for separate image acquisitions tailored to a pixel-dependent cardiac coordinate system, which requires additional computational resources and introduces estimation errors due to probe orientation variability. Furthermore, most systems omit shear strain information due to the complexity of displaying all components clearly. To overcome these challenges, a new study introduces an innovative 3-D high-spatial-resolution, coordinate-independent strain imaging approach that utilizes principal stretch and axis estimation. This method transforms all strain components into three principal stretches along three orthogonal principal axes, facilitating direct visualization of primary deformations and simplifying interpretation.
The proposed methodology uses an advanced 3-D speckle tracking technique that includes tilt filtering. It incorporates randomized searching within a two-pass tracking framework and employs phase rotation of the 3-D correlation function for enhanced filtering accuracy. These improvements significantly refine displacement gradient estimation, especially for those related to the axial displacement component. Non-axial displacement gradients are also more accurately estimated through a correlation-weighted least-squares method, constrained by tissue incompressibility.
Testing with simulated and in vivo canine cardiac datasets to estimate Lagrangian strains from end-diastole to end-systole showed that the proposed method reduces displacement estimation errors by factors of 4.3 to 10.5 over traditional one-pass processing. Moreover, maximum principal axis/direction imaging has proven more effective in detecting localized disease regions than conventional strain imaging techniques. In summary, this study presents a more accurate and robust method for strain imaging, offering significant improvements in the detection of myocardial abnormalities compared to existing techniques.