Plenary Talks
Engineering Hope with Biomimetic Microelectronic Systems
Dr. Wentai Liu
University of California at Santa Cruz
Research in biomimetics has progressed rapidly in the recent years fueled by the unique interdisciplinary efforts fusing engineering, medicine, and biology. This interdisciplinary research requires addressing the aspects of humanity and societal impacts, technical challenges and barriers, targeting a wide range of applications. These applications span from understanding the highly complex biological systems, to treating/restoring/repairing the lost biological functions such as deafness, blindness, and paralysis, to building human-machine interface for performance enhancement (super person). Biomimetic systems will offer viable solutions to neuro-disorder diseases which potentially affect very large population of people worldwide and thus occupy largest market share in healthcare.
The major enabling technology for implantable biomimetic systems include biological recording, stimulation, bio-signal processing, wireless communication, sensing, electrode, hermetic packaging, and powering, where the implants must deal with critical constraints of size, power, reliability, safety, and technology. The additional heterogeneous system testing/measurements under the regulatory and compliance guidelines critically differentiate from the conventional electronic system designs and accordingly require new design methodology at every design level. Clearly, integration and miniaturization of the implants become very essential and require solutions from many fronts – device, circuit, architecture, system, algorithm, design, testing, packaging, and technology.
This talk defines and presents enabling technologies and design challenges to realize integrated and miniaturized biomimetic systems, especially neural implants, which can be used for building advanced neuroscience and neuroprosthetics platform with closed-loop control mechanisms. Works on neural implants consist of four major functional blocks of recording, stimulation, processing, and wireless communication, each of which has different challenges to be overcome. Seamless on-chip integration of these fundamental blocks with low power requirements inevitably needs mixed-voltage and mixed signal design techniques. Combining these techniques with stated-of-art high voltage process, high compliance voltage for integrated nerve/neuromuscular stimulators becomes possible. Fundamental understanding of the noise sources and their contributions within neural interface provides a better way to optimize the design of the recording functions and thus improves the performance of the neural data acquisition and processing modules. Very high bandwidth transceivers with efficient protocols enable the implant to support multiple biological objects without constraints from the data bandwidth. Sophisticated bio-signal processing techniques and their integration/miniaturization for tasks such as real-time spike sorting, classification, cognition recognition, etc. are critical to decision-making and closed-loop execution in future applications. This talk will cover the development history, market opportunity, technical challenge/barriers, enabling technology, and application examples.