A Robust Method to Estimate the Time Constant of Elastographic Parameters.

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Computational Mechanics

Background:

• Viscoelastic and poroelastic elastography require accurate estimation of temporal strain behavior.
• Traditional curve-fitting methods (LM, NM, TR) for time constant (TC) estimation can be inaccurate and sensitive to initial guesses and noise.

Purpose of the Study:

• To introduce and validate a novel variable projection (VP) technique for robust and accurate estimation of TC and steady-state values in elastography.
• To compare the performance of VP against established curve-fitting methods using simulations and experimental data.

Main Methods:

• Development of a variable projection (VP) technique for temporal curve analysis.
• Application of VP with a novel analytical model applicable to both poroelastic and viscoelastic tissues.
• Validation using finite element simulations, ultrasound simulations, and experimental data.

Main Results:

• VP demonstrates superior accuracy and robustness to noise compared to LM, NM, and TR methods in TC estimation.
• VP's performance is independent of initial TC guess accuracy, handling a wide range of values.
• Experimental results show VP reliably estimates axial strain TC, while traditional methods frequently fail to converge or yield incorrect solutions.

Conclusions:

• The variable projection (VP) technique offers a more accurate and reliable approach for estimating time constants in elastography.
• VP provides a robust solution for analyzing temporal elastographic data, particularly in the presence of noise and with variable initial conditions.
• VP enhances the clinical applicability of viscoelastic and poroelastic elastography by improving parameter estimation accuracy.