作者
Kai Zhang,Zhengxun Song,Rongqiang Wei,Shiwei Zhang,Liguo Tian,Zuobin Wang
摘要
Label-free electromagnetic identification of breast cancer cells represents a promising research frontier in biomedical detection. Conventional cellular detection techniques typically depend on fluorescent labeling or morphological imaging, yet they fail to directly capture disparities in intracellular dielectric properties. Furthermore, systematic simulation frameworks are still absent for optimal frequency band selection and characteristic parameter extraction in near-field cellular detection using microwave atomic force microscopy. In this work, a microwave atomic force microscopy near-field detection model was established in COMSOL to simulate the electromagnetic responses of MCF-10A normal breast epithelial cells, MCF-7 low-malignancy breast cancer cells, and MDA-MB-231 high-malignancy breast cancer cells over 1-20 GHz. The three cell models used the same probe structure, port excitation, and boundary conditions, while their geometric dimensions, relative permittivity, and conductivity were specified according to the corresponding cell types. A frequency step of 0.1 GHz was used, yielding port-response data at 191 frequency points. The results show that all three cell models are under strong reflection, making direct comparison of$|\mathrm{S}_{11}|$ineffective for cell-type discrimination. In contrast, complex$\mathrm{S}_{11}$separation, equivalent probe admittance, and equivalent capacitance are more sensitive to electromagnetic differences among the cell models. The maximum complex$\mathrm{S}_{11}$separation between MCF-7 and MCF-10A occurs at 13.9 GHz, whereas the maximum$|{\Delta} \mathrm{S}_{11}|$values for MDA-MB-231 versus MCF-10A and MDA-MB-231 versus MCF-7 occur at 20.0 GHz. At 20 GHz, the equivalent admittance and equivalent capacitance of MDA-MB-231 increase by approximately 4.00% and 3.80%, respectively, relative to the normal-cell model, indicating stronger conductive-loss and capacitive-coupling responses. Overall, frequencies around 13.9 GHz are optimal for discriminating MCF -7 cells from normal mammary cells, while the 16-20 GHz band, particularly the frequency range close to 20 GHz, is better suited to identify the highly malignant MDA-MB-231 cell model. This work establishes a numerical simulation foundation for label-free electromagnetic discrimination and microwave imaging of breast cancer cells.