Brain Computer Interfaces: an fNIR research study

November 26, 2007
2:50 pm - 4:00 pm
Halligan 111
Speaker: Audrey Girouard, Tufts Computer Science - Research Talk


Functional Near-Infrared Spectroscopy (fNIR) is a new brain imagery tool showing potential for use in the field of human computer interaction (HCI) because of its lightweight, non-invasive qualities. fNIR could become an additional input to interfaces, by recoding the user's mental state through the measure of blood flow. The talk will be composed of two parts: a report on the feasibility study, and a discussion of my plans for future research. We performed a feasibility study to establish whether we could measure metal workload. A simple and controlled experiment used colored cubed to simulate different levels of mental workload. Results show the ability to differentiate between levels of workload, by using machine learning algorithms to classify them. My plans for future work use emotions with fNIR. First, a short-term study measuring emotional arousal will test fNIR' potential to measure emotion. Given positive results with this study, I intend to use fNIR in communication interfaces, such as email and chat. Measuring the emotions of the user while using such interfaces should prove useful, leading to adapting these communication interfaces, or other interfaces the user is currently interacting with.