Development of Microfluidic Impedance Flow Cytometry Enabling the Quantification of Intrinsic Single-Cell Bio-Dielectric Parameters Based on Constrictional Microchannel, Numerical Simulation, and Neural Network

This study presents a microfluidic impedance flow cytometry enabling the quantification of intrinsic single-cell bio-dielectric parameters (e.g., cell radius of
${R}_{\text {c}}$
, relative membrane permittivity of
$\varepsilon _{\text {mem}}$
, and cytoplasmic conductivity of
$\sigma _{\text {cyto}}$
) based on constrictional microchannel, numerical simulation, and neural network. Numerical simulation of a traveling cell within constrictional microchannel was conducted to locate relationship between intrinsic bio-dielectric parameters of single cells with corresponding impedance profiles, functioning as “reference” cells for the feedforward neural network, producing fitting accuracies of 1.00 ± 0.00 of
${R}_{\text {c}}$
, 1.00 ± 0.00 of
$\varepsilon _{\text {mem}}$
, and 1.00 ± 0.01 of
$\sigma _{\text {cyto}}$
. Then, the fitted analytical equation based on numerical simulation and neural network was further used to translate measured impedance profiles into intrinsic bio-dielectric parameters of single cells, where
${R}_{\text {c}}$
,
$\varepsilon _{\text {mem}}$
, and
$\sigma _{\text {cyto}}$
were quantified as 7.6 ±
$0.8 \mu \text{m}$
, 16.73 ± 2.97, and 1.38 ± 0.34 S/m of K562 cells (
${n}_{\text {cell}}=1284$
), 6.4 ±
$0.8 \mu \text{m}$
, 17.47 ± 5.09, and 1.46 ± 0.47 S/m of Jurkat cells (
${n}_{\text {cell}}=2116$
), 5.9 ±
$0.7 \mu \text{m}$
, 17.40 ± 3.14, and 1.96 ± 0.81 S/m of HL-60 cells (
${n}_{\text {cell}}=3543$
), respectively. Compared to phenomenal electrical signals (i.e., real and imaginary parts at specific frequency) detected by conventional electrical flow cytometry, which were prone to environmental variations, this microfluidic platform enabled the quantification of intrinsic bio-dielectric parameters of single cells and thus provides an effective tool in the field of flow cytometry.