Temiloluwa Olubanjo and Hakan Toreyin won two of the top three student paper competition awards at the 2014 IEEE Engineering in Medicine and Biology Conference, held August 26-30 in Chicago. Both students in the Georgia Tech School of Electrical and Computer Engineering (ECE), Olubanjo placed first in the competition, while Toreyin placed third and was named the North American finalist in the contest. 

Co-advised by ECE Associate Professors Maysam Ghovanloo and Elliot Moore, Olubanjo was honored for her paper entitled “Tracheal Activity Recognition Based on Acoustic Signals.” Tracheal activities detected by a neck-worn device are among signals that, if recognized properly, can contribute to effective health monitoring. In this paper co-written with Ghovanloo, Olubanjo explored novel signal processing algorithms to detect and classify common tracheal events that can provide insight into an individual’s health and well-being, such as breathing, chewing, swallowing, coughing, clearing the throat, and speech. Monitoring the patients’ compliance with their prescribed medication regimens and estimating the energy balance in the obese and overweight population are among other applications of a tracheal activity recognition system that is currently under development in the GT-Bionics Lab, which is led by Ghovanloo.

Toreyin’s paper was entitled “A Low-Power, Time-Division-Multiplexed Vector Matrix-Multiplier for a Vestibular Prosthesis,” and his coauthor was his Ph.D. advisor Pamela Bhatti. A vestibular prosthesis can potentially improve the quality-of-life for persons suffering from balance or vision problems caused by bilateral vestibular dysfunction, disorders associated with the brain and inner ear. By replacing the function of peripheral vestibular organs, the prosthesis senses 3D angular and linear head motions by using inertial sensors and selectively stimulates the corresponding vestibular neurons. Toreyin’s paper reports on the first integrated circuit (IC), known as a vector matrix multiplier, designed to improve the effectiveness of stimulating vestibular neurons. Operating in the CMOS subthreshold region, the IC performs a 3-by-3 vector matrix multiplication of rate sensor outputs to reduce misalignment between implanted angular rate sensors and associated peripheral sense organs, and provides a means to precompensate for unwanted electrical stimulation of vestibular neurons.