Microwave Flow Cytometer for Single Cell Detection, Differentiation, and Drug Effect Study

New methods to rapidly detect and identify pathogens in patients’ blood are needed for timely disease treatment. In this work, a tunable microwave interferometer was used to measure single Candida cells of C. albicans, C. tropicalis, C. parapsilosis, and C. krusei at multiple frequencies between 0.265 GHz and 7.76 GHz. A Quadratic Discriminate Analysis was used to classify Candida species with an accuracy of 0.875. Likewise, a microwave device was used to measure single Escherichia coli cells of B and K-12 strains at multiple frequencies between 500 MHz and 8 GHz. A LightGBM model was developed to classify cell types at a high accuracy of 0.96 at 1 GHz. Permittivity values Δε’ and Δε” of non-budding single Candida were colinear with cell volumes and had a coefficient that depends on measurement frequency, cell species and viability. Viable and non-viable cells had significant permittivity differences at lower frequency spectrum, but with substantial overlap at 7.76 GHz. Cell heterogeneity characterization for C. tropicalis and C. albicans was studied. The obtained scatter plots of cell permittivity for both species show a colinear relationship for cells of different volumes, which indicates a largely conserved cell molecular composition profile despite heterogeneous cell sizes. It was determined that treating Candida cells with Caspofungin (CSP) diacetate in dimethyl sulfoxide (DMSO) for 10 minutes significantly altered cell microwave properties. This change was likely due to damaged cell wall and membranes, and the level of change was Candida species dependent. Additionally, it was observed that drugs induce significant heterogeneity increase. The results show that microwave flow cytometry is a promising new method for high throughput and noninvasive characterization of cell heterogeneity. Moreover, broadband microwave measurement is a promising new approach for physicians to provide personalized candidemia therapy. Lastly, a system consisting of a microfluidic mixer, separator and microwave device was used to separate and measure Candida cells in spiked bovine blood within 2 hours. The results showed that microwave sensors are promising to extract, detect and differentiate Candida cells from infected blood.