Application of Heuristic Optimization in Bioimpedance Spectroscopy Evaluation

As material analysis method, the bioimpedance spectroscopy (BIS) is an effective technique to obtain material characteristics in many biological applications. Among the BIS models, the single-dispersion RC model and the Cole-Cole model constitute the fundamental mathematical descriptions for BIS measurements. These models enable both the evaluation of impedance spectrum and derivation of core parameters that describe biological processes. This paper presents the application of the particle swarm optimization (PSO), as a flexible heuristic optimization approach, to both derive the model parameters and characterize biological media based on BIS measurements. First, the fundamental modeling approaches are addressed and the core parameters are discussed. Moreover, the employed electrical impedance measuring instrument and conducted experiments are presented. Then, a multi-objective fitness function is established for the usage of the PSO algorithm for model parameter optimization. The paper demonstrates two case studies, namely, BIS measurements are performed to monitor i) liver fat and characterize its state with an optimized Cole-Cole model and ii) cell culture growth and characterize its state with an optimized single-dispersion RC model. It is shown with experimental results that PSO is an effective and robust tool to fit these biological models to BIS measurements. Additionally, the experimental results also highlight that the cell culture growth process cannot be modeled properly with the single-dispersion RC model in high frequency ranges. This experimental observation establishes the demand for the derivation of a more sophisticated mathematical model for the comprehensive characterization of cell culture growth in wide frequency spectrum.