Interpretation of preliminary electrical data in impedance flow cytometry: numerical simulation, theoretical analysis, and neural net fitting
Xiao Chen,Minruihong Wang,Yan Liu,Mingyue Liu,Deyong Chen,Bo Chen,Junbo Wang,Jian Chen
DOI: https://doi.org/10.1007/s10404-022-02615-2
2023-01-04
Microfluidics and Nanofluidics
Abstract:As a label-free tool, single-cell impedance flow cytometry is an effective method to measure electrical properties and distinguish cell populations. However, interpretation of preliminary electrical data in impedance flow cytometry into single-cell intrinsic electrical properties (e.g., membrane relative permittivity of ε mem and cytoplasmic conductivity of σ cyto ) is full of challenges. In this study, using a multiple-frequency Coulter counter as the research model, numerical simulation was first conducted to locate relationship between ε mem /σ cyto and multiple-frequency impedance values, which were used as inputting data for the Maxwell's mixture equation to obtain ε mem /σ cyto theoretically for comparison. Furthermore, neural network was adopted to fit multiple-frequency impedance values with numerical simulation, and fitted analytical equations were further used to obtain ε mem /σ cyto for comparison. Relatively low vs. high accuracies in quantifying single-cell ε mem and σ cyto were found based on the Maxwell's mixture equation vs. neural net fitting, where values of ε mem @Ma/ε mem @si vs. ε mem @ne/ε mem @si and σ cyto @Ma/σ cyto @si vs. σ cyto @ne/σ cyto @si were quantified as 0.857 ± 0.189 vs. 0.99997 ± 0.00058 and 0.947 ± 0.108 vs. 1.00024 ± 0.01025, respectively. Thus, neural net fitting can function as an effective approach in interpreting preliminary electrical data in impedance flow cytometry.
physics, fluids & plasmas,nanoscience & nanotechnology,instruments & instrumentation