A novel discrete particle swarm optimization algorithm for estimating dielectric constants of tissue

Area of Science:

• Electromagnetics and Computational Physics
• Biomedical Imaging and Signal Processing

Background:

• Global optimization algorithms iteratively refine solutions to complex problems.
• Accurate modeling of biological tissues is crucial for medical applications, particularly in microwave imaging.
• Existing methods for permittivity estimation in multilayered lossy structures can be time-consuming.

Purpose of the Study:

• To introduce a novel discrete particle swarm optimization (PSO) algorithm for permittivity estimation.
• To demonstrate the efficiency and flexibility of PSO-based methods in complex biomedical imaging scenarios.
• To analyze the performance of the proposed algorithm in reconstructing images using microwave imaging (MI).

Main Methods:

• Development of a discrete particle swarm optimization algorithm tailored for permittivity estimation.
• Application of the algorithm to reconstruct images of lossy multilayered body tissue models using microwave imaging.
• Systematic investigation of parameter impacts (MI frequency, immersion medium, agent count, smoothing coefficient, max velocity) on estimation error.

Main Results:

• The proposed discrete PSO algorithm significantly improves estimation time by 85% compared to previous methods.
• Accurate permittivity estimations with a maximum error below 10% were achieved in as little as 0.1 minutes.
• Parameter sensitivity analysis identified optimal settings for enhanced estimation performance.

Conclusions:

• The developed discrete PSO algorithm offers a highly efficient and accurate solution for permittivity estimation in biomedical microwave imaging.
• PSO-based approaches are flexible and effective for handling complex inverse problems in imaging.
• Optimized parameter selection is critical for achieving rapid and precise tissue property reconstruction.