Evaluation of inverse methods for estimation of mechanical parameters in solid tumors

Area of Science:

• Biomedical Engineering
• Computational Biology
• Cancer Research

Background:

• Accurate estimation of mechanical parameters in living tissues is crucial for effective cancer treatment.
• Understanding tissue hydraulic conductivity (k) and hydraulic conductivity ratio (α) aids in developing targeted therapies.

Purpose of the Study:

• To develop and validate an inverse method for estimating hydraulic conductivity (k) and hydraulic conductivity ratio (α) in living tissues.
• To compare the performance of different inverse algorithms for parameter estimation.
• To assess the impact of measurement location on estimation accuracy.

Main Methods:

• Comparative analysis of population-based (Levenberg-Marquardt, conjugate gradient) and gradient-based (genetic algorithm, particle swarm optimization) inverse methods.
• Validation of robust algorithms (LM, GA, PSO) using experimental data.
• Investigation of single-point interstitial pressure measurement effects on hydraulic conductivity estimation.

Main Results:

• Levenberg-Marquardt, genetic algorithm, and particle swarm optimization methods demonstrated convergence with <10% estimation error in both noise-free and noisy data.
• Conjugate gradient method failed to converge with noisy data.
• Experimental validation confirmed the accuracy of the LM, GA, and PSO algorithms for estimating k and α in mammary adenocarcinoma (R3230AC) tissue.
• Estimation errors were higher for measurements taken near the tumor's surface and center.

Conclusions:

• The developed inverse method using LM, GA, and PSO algorithms is effective for estimating hydraulic conductivity and ratio in living tumor tissues.
• Measurement location significantly impacts the accuracy of hydraulic conductivity estimation, with central and surface points yielding greater errors.
• This approach offers a valuable tool for cancer research and treatment planning by providing essential mechanical tissue parameters.