A microfluidic system for dielectrophoretic characterization of cancer cells

Sel, Kaan
Dielectrophoresis (DEP) is a promising cell manipulation approach for early diagnosis of cancer, which significantly increases chances of successful treatment. Compared to other cell manipulation techniques that rely on surface antigens, DEP systems enable label-free, cost-effective, simply-implementable cell characterization and separation. However, separation efficiency of the DEP based systems is limited and still far from meeting the medical requirements for early cancer detection. In order to improve the throughput of current DEP systems, it is important to obtain the optimum operating conditions. The main objective of this thesis is to conduct accurate dielectrophoretic characterization of cancer cells without ascertaining cell dielectric properties at different operating conditions in autonomous fashion. The presented system integrates a microfluidic DEP device with a CMOS image sensor, and a portable signal generator. The system enables DEP spectra analysis of cells in a wide frequency band (30 kHz to 50 MHz). The microfluidic DEP device, contains optimized electrode structures that can generate isomotive electric-field inside the analysis region. Hence, cell motion under the DEP force can directly be related to its dielectrophoretic behaviour. In addition, post-processing can be done either in a custom-developed MATLAB GUI or in a custom-developed Android software connected to a smartphone using an automated cell tracking algorithm. With the DEP characterization device presented in this thesis, different conditions (operating frequency, medium characteristics) can be tested in a portable, autonomous and rigorous fashion to find the optimum case for cell separation. The system was tested with both MFC-7 (Human Breast Adenocarcinoma) and K562 (Human Chronic Myeloid Leukemia) cells due to availability. The results display consistency with the DEP spectrum studies conducted with these two cell groups in the literature.