Xiaolin Li, Yini Lian, Xu Bai, Peng Suo, Lijun XU, Jiangtao Sun
Measuring and recognizing single cells are fundamental challenges in biology. Electrical impedance tomography (EIT) is a non-destructive and label-free visualization monitoring technique, but its application in single-cell monitoring is limited by high contact resistance at the microscale and complex fabrication challenges. In this work, an easily fabricable EIT sensor at microscale is proposed, in which the liquid-electrode design enables a sufficiently large contact area between metal lead and liquid and thus significantly decreases the contact impedance, allowing the imaging of the shape, position and size of a single cell or multiple cells. The genetic algorithm is employed to optimize the sensor structure, resulting in a sensitive and concentrated sensing region at the center of the sensor to eliminate interferences from the outside surroundings. To better image cells, an adaptive block sparse Bayesian learning algorithm for small targets reconstruction is adopted and proven effective in removing artifacts near cells and at the edges of the sensitive region. Finally, experiments to image PS microspheres and yeast cells at microscale were conducted as a proof of concept. The result shows that the proposed sensor with liquid electrodes can effectively reconstruct the shapes, positions, and sizes of microspheres and cells, where the estimation errors of the microsphere size are less than 5%. The proposed liquid-electrodes EIT sensor provides a low-cost promising alternative to image single cell at microscale, which can enable the morphological characterization of single cell.