Comparison of shift estimation strategies in spectral elastography

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Signal Processing

Background:

• Spectral elastography is a technique used to measure tissue stiffness.
• Accurate spectral shift estimation is crucial for reliable elastographic measurements.
• Existing methods like normalized cross-correlation (NCC) are computationally intensive.

Purpose of the Study:

• To compare the performance of different spectral shift estimators in spectral elastography.
• To identify computationally efficient estimators without compromising accuracy.
• To evaluate the impact of subsample estimation techniques on performance.

Main Methods:

• Evaluated normalized cross-correlation (NCC), sum squared difference (SSD), and sum absolute difference (SAD) spectral estimators.
• Utilized simulation and experimental data for performance assessment.
• Investigated cosine-fit and parabolic-fit interpolation methods for subsample estimation.

Main Results:

• Spectral SSD-based elastography showed performance comparable to NCC but with lower computational cost.
• Spectral SAD-based strain estimation was less accurate than NCC and SSD.
• Parabolic-fit subsample estimation was analyzed for computational simplicity, despite cosine-fit offering better bias and standard deviation reduction.

Conclusions:

• The spectral SSD strain estimator, combined with parabolic-based subsample estimation, is recommended for spectral elastography due to its balance of performance and computational efficiency.
• This approach offers a practical alternative to more computationally demanding methods like NCC.