Discrimination of Tumor Cell Type Based on Cytometric Detection of Dielectric Properties

Dezhi Tang,Lin Jiang,Nan Xiang,Zhonghua Ni
DOI: https://doi.org/10.1016/j.talanta.2022.123524
IF: 6.1
2022-01-01
Talanta
Abstract:In this paper, a microfluidic impedance cytometer (MIC) was employed to analyze the dielectric properties of human white blood cells (WBCs) and four tumor cell lines and realize the label-free identification of cell types. The impedance of cells was detected using an asymmetric serpentine microchannel based MIC under four different frequencies simultaneously. The asymmetric serpentine microchannel achieved the elasto-inertial focusing of cells into a single train, ensuring accurate impedance detection of cells. Various dielectric parameters (cell diameters, impedance amplitude |Z|, impedance phase shift ΦZ, and electric opacities |Z|HF/|Z|LF, ΦZHF/ΦZLF, Re(ZHF)/Re(ZLF), and Im(ZHF)/Im(ZLF)) were defined and used to analyze the dielectric properties of cells. The obtained dielectric parameters were used to train machine learning classification models for identifying cell types. Using all parameters proposed in this paper (cell diameter, opacity |Z|HF/|Z|LF, ΦZHF/ΦZLF, Re(ZHF)/Re(ZLF), and Im(ZHF)/Im(ZLF)) to train the classification model, the true positive rate (TPR) for the identification of WBCs, A549, MCF7, H226, and H460 cells were 99.6%, 96.2%, 99.1%, 97.6%, and 97.2%, respectively. Results showed that our MIC provided a promising method for label-free discrimination of circulating tumor cells in multiple primary cancers.
What problem does this paper attempt to address?