PhD Thesis Defense: Jenny Qiu

Friday, July 26, 2019, 9:00–11:00am

Jackson Conference Rm, Cummings Hall

“Measuring brain activity with a wearable, low-cost electroencephalograph system in interactive experimental paradigms”

Abstract

The development of smaller and power efficient electronics over the last few decades has facilitated the growth of lower-cost, wearable versions of common medical devices. As a result, low-cost, wearable systems that measure neural activity using electroencephalography (EEG) — a non-invasive technique that records electrical activity from the brain — have entered the consumer marketplace. From this expansion, there is a need to scientifically assess the potential applications and capabilities of these wearables to measure reliable neural signals.

The thesis presents a low-cost version of a $100k traditional system. The optimized wearable system, which includes the low-cost OpenBCI Cyton Board with Daisy chain EEG amplifier, is commercially available at $1.5k. The thesis first describes the optimization of the out-of-box version of the OpenBCI Cyton Board. The Cyton Board had a number of deficiencies such as timing discrepancies that needed to be addressed. Several software patches and hardware modules were developed and implemented. The second part of the thesis details two experiments conducted to benchmark the optimized EEG system using contemporary studies of neural signals measured with clinical EEG systems. The first experiment evaluated neural data using an oddball paradigm and the second experiment evaluated neural data as elicited by a novel interactive version of a classic paradigm. These experiments demonstrate the potential of wearable EEG for collecting high-quality neural signals in everyday workspaces. The third part of the thesis introduces a support vector machine classification system that is then applied to data from the interactive benchmark experiment. The classifier receives features of a single-trial event-related potential (ERP) and predicts the interactive user feedback that initiated the ERP.

The presented body of work aims to provide researchers with accessible and affordable neural measurement techniques outside of traditional systems and to expand the computer environments in which these neural signals can be monitored.

Thesis Committee

For more information, contact Daryl Laware at daryl.a.laware@dartmouth.edu.