A CMOS 22k-pixel single-cell resolution multi-modality real-time cellular sensing array

Abstract

This paper presents a single-cell-resolution multimodality cellular interfacing array for holistic cell characterization and high-throughput drug screening. It contains 22,528 multimodality sensing pixels with a pixel size of 17μm×17μm. Each pixel contains a 4μm×4μm gold-plated electrode for extracellular potential recording and a 3μm×6μm photodiode for optical detection. To compensate for the degraded signal to noise ratio (SNR) due to the large electrode-electrolyte impedance at the cell-sensor interface, we employ a DC-coupled chopper amplifier when performing extracellular potential recording. A proof-of-concept design is implemented in a standard 130nm CMOS process. It achieves a measured input-referred noise of 2.46μVrms integrated from 1Hz to 4kHz with a chopping frequency of 20kHz, and cell-based measurement results demonstrate real-time recorded action potentials with high SNR. Measured high-resolution optical shadow images provide the detailed cell and tissue morphology as well as the 2D distribution of on-chip cultured cells. Furthermore, this paper presents the first CMOS sensing array to successfully capture on-chip rat cardiomyocyte contraction using multiple sensing modalities, specifically optical detection and potential recording.