Ultra-Fast On-Chip Sensing for PCR Detection with Machine Learning-Based Image Processing Algorithm

Abstract

This work developed a highly integrated and ultra- fast optoelectronic on-chip Polymerase Chain Reaction (PCR) system, in which multiple Si-based optoelectronic on-chip sensing techniques have been used, i.e., CMOS imaging sensor, novel snake-shaped microfluidic chip fabricated using MEMS technology, and a thermoelectric cooler chip module. A machine learning-based algorithm has also been developed to improve the PCR testing speed and accuracy. This design exhibits an ultra-fast PCR testing process of fewer than 5 minutes, and a PCR testing accuracy as high as 91.1 % has been achieved.