Decoding Subjective Preference From Single-Trial Near-Infrared Spectroscopy Signals
By Luu, Sheena; Chau, Tom; Journal of Neural Engineering, Vol. 6, No. 1, pp. 1-8Publication Date: 2009
Study evaluated the effectiveness of a non-invasive brain-computer interface using near-infrared spectroscopy (NIRS-BCI) for indicating user preference. The study was conducted as part of an ultimate goal of developing technology for people with severe disabilities such as locked-in syndrome. For the study, 9 young adult participants were fitted with headgear, worn on the forehead, on which a multichannel frequency domain NIRS instrument was mounted, emitting light into the pre-frontal cortex of the brain. Participants were seated in front of a computer monitor displaying images of two drinks and were instructed to mentally decide which of the drinks they preferred. They were then asked to indicate their preferred choice by clicking with a mouse on a “First” and “Second” button on the computer screen. NIRS recordings of participants’ prefrontal cortex were taken during drink presentation but not while participants used the mouse to indicate their preference. A total of 56 trials were used for each participant. Using simple features and a linear classification scheme, the researchers were able to decode which drink was preferred with an average accuracy of 80 percent. Study limitations and implications for future research are discussed.
Published by: IOP Publishing Ltd (Website:http://www.iop.org/)
Link to text: http://www.iop.org/EJ/abstract/1741-2552/6/1/016003/
