The talk will present methods for practical realization of sophisticated digital signal processing (DSP) algorithms for healthcare applications that require many-channel electrophysiological recordings. An automated architecture design framework will generate dual output: real-time solution that meets power density constraints of medical implants, and solution for accelerated processing by hardware emulation. Orders of magnitude improvements in the number of recording channels and decreased hardware cost will be demonstrated. A real-time implantable DSP chip will demonstrate simultaneous processing of 64 channels, with over 90% data compression for wireless telemetry. A DSP architecture for hardware emulation can achieve a 10,000 times speed-up in data processing compared to state-of-the-art computers. A successful integration of neural-data processing will significantly advance many applications such as visual, auditory, motor, and cognitive prosthetics. Methods for achieving faster analysis of electrophysiological data will provide neuroscientists quicker access to important research data and improve the overall quality of living for persons with neurological disorders. The methods presented here are also applicable to other emerging biomedical applications that require energy- and cost-efficient data processing.
Dejan Marković is an Assistant Professor of Electrical Engineering at the University of California, Los Angeles; and a member of the UCLA Biomedical Engineering Interdepartmental Program. He completed the Ph.D. degree in 2006 at the University of California, Berkeley. In recognition of the impact of his Ph.D. work, he was awarded 2007 David J. Sakrison Memorial Prize at UC Berkeley. His current research is focused on integrated circuits for emerging radio and healthcare systems, design with post-CMOS devices, optimization methods and CAD flows. He received an NSF CAREER Award in 2009 for the research described in this talk.