System-on-Chip for Flow Cytometry With Impedance Measurement and Integrated Real-Time Size Classification

This paper presents an impedance measurement system-on-chip (SoC) for flow cytometry (i.e. cell counting) applications. A source-differential, three-electrode sensing scheme is used in a microfluidic flow cell for particle detection. At the front-end, a lock-in amplifier architecture is used, including a high-gain TIA with 60 MHz bandwidth, passive mixers, and low-pass filters. The ac sensor signal is demodulated to extract in-phase (I) and quadrature (Q) baseband components to measure complex impedance. At the back-end, the SoC includes an 8-bit level-crossing ADC (LCADC) for digitizing I/Q signals, followed by real-time digital feature extraction and linear classification for real-time cell size determination. The SoC was fabricated in a 180 nm CMOS process. A measured prototype IC achieves 733 fA/$\sqrt{Hz}$ noise floor and 23 pArms input-referred noise from 1-1 kHz. Combined with a microfluidic flow cell, polymer beads in solution were used as cell surrogates to demonstrate particle counting. Measured results for particle diameters of 10 $\mu$m, 6 $\mu$m, 4.5 $\mu$m and 3 $\mu$m are shown. Following offline training, the SoC demonstrated on-chip classification of 4.5 $\mu$m and 6 $\mu$m beads with a prediction accuracy of 86.16% with pre-recorded data, and 73.6 % while performing real-time inline classification.