ESE Spring Colloquium – “Autonomous Microsystems Based on Heterogeneously Integrated CMOS for Biological Big Data”

Minimally invasive and chronic physiological monitoring can provide an effective means of disease prevention and early detection while the cumulative big data can unveil hidden patterns in our physiology. Yet, current physiological monitoring tools are often bulky, invasive, and expensive, limiting their sensitivity and applicability. In this talk, I will discuss autonomous microsystems based on heterogeneously integrated CMOS, a platform on which ideal physiological sensors and actuators can be built.

A micro-scale optoelectronically transduced electrode (MOTE), an exemplary microsystem I have designed and built for tetherless neural recording, is powered and communicates optically through a vertically integrated AlGaAs micro-scale light emitting diode (µLED), eliminating the needs for a battery or a RF coil; the MOTE is smaller than a human hair (~60 µm × 30 µm × 330 µm) and weighs about one 1 µg (cf. a grain of sand is about 670 µg). I will review the unique challenges and considerations in developing such heterogeneous systems in terms of device fabrication, circuit design, integration, and handling/manipulation.

While the MOTE is designed for neural recording, its design methodologies can also be used to monitor other physiological parameters such as temperature, pH, glucose-level, etc. I will introduce future autonomous microsystems with expanded modalities and how to interface them with existing wearables. As such microsystems become more accessible, the resulting biological big data will help enable personalized healthcare and produce a physiological ‘digital twin’ (like the architectural digital twins of select cities) that can add a new dimension to epidemiological and aging studies.