Presentation Title

Distinctness of EEG Based Brain Signal Readings and Their Potential for Biometric Systems

Location

Nessmith-Lane Atrium

Session Format

Poster Presentation

Research Area Topic:

Engineering and Material Sciences - Mechanical

Abstract

A human brain consists of a large concentration of neurons which create minute currents while communicating with each other. During specific tasks such as closing and opening of eyes or movement of the tongue, these currents are recorded using an electroencephalogram (EEG) noninvasively through placing electrodes on the scalp of a person. There have been avid studies on signals acquired through the EEG to physically move objects using machines and prosthetics. In these studies, it has been stated that the brain signals have been distinct based on the subject (person) but there hasn’t been any research conducted to show that brain signals are truly distinct and can be considered as a strong biometric signature. The motivation for this study was drawn from contrastive results acquired for multiple subjects during a previous research conducted on finding the most effective electrodes for a brain-computer interface (BCI). This study will use the publicly available recorded, refined, and filtered data for nine subjects from an online database (BCI competition IV Ð Dataset 2a and 2b). The data consists of signals acquired from 22 electrodes placed at specific locations on the scalp and sampled at a rate of 256 Hz. The process of placing electrodes and recording data is performed during three separate sessions. Among the three sessions acquired from the same subject, a strong correlation coefficient is sought to proceed to the next stage. Here the correlation coefficient is the measure of similarity between two signals and the similarity can range from Ô0Õ to Ô1Õ. If and when a consistent correlation of 0.6 or more is acquired out of a maximum of Ô1Õ for the data in all three sessions, then the research will proceed to compare signal data from different subjects for a relatively lower correlation coefficient. Data acquired in preliminary simulations show a distinctive contrast between subjects and in this study the primary focus will be finding a strong similarity among the brain signals acquired for the same subject. So in conclusion each subject’s data will be correlated for the same task (eyes opened and eyes closed) across three separate sessions then the one subject’s data will be correlated with another and the findings will be presented in the symposium.

Presentation Type and Release Option

Presentation (Open Access)

Start Date

4-16-2016 2:45 PM

End Date

4-16-2016 4:00 PM

This document is currently not available here.

Share

COinS
 
Apr 16th, 2:45 PM Apr 16th, 4:00 PM

Distinctness of EEG Based Brain Signal Readings and Their Potential for Biometric Systems

Nessmith-Lane Atrium

A human brain consists of a large concentration of neurons which create minute currents while communicating with each other. During specific tasks such as closing and opening of eyes or movement of the tongue, these currents are recorded using an electroencephalogram (EEG) noninvasively through placing electrodes on the scalp of a person. There have been avid studies on signals acquired through the EEG to physically move objects using machines and prosthetics. In these studies, it has been stated that the brain signals have been distinct based on the subject (person) but there hasn’t been any research conducted to show that brain signals are truly distinct and can be considered as a strong biometric signature. The motivation for this study was drawn from contrastive results acquired for multiple subjects during a previous research conducted on finding the most effective electrodes for a brain-computer interface (BCI). This study will use the publicly available recorded, refined, and filtered data for nine subjects from an online database (BCI competition IV Ð Dataset 2a and 2b). The data consists of signals acquired from 22 electrodes placed at specific locations on the scalp and sampled at a rate of 256 Hz. The process of placing electrodes and recording data is performed during three separate sessions. Among the three sessions acquired from the same subject, a strong correlation coefficient is sought to proceed to the next stage. Here the correlation coefficient is the measure of similarity between two signals and the similarity can range from Ô0Õ to Ô1Õ. If and when a consistent correlation of 0.6 or more is acquired out of a maximum of Ô1Õ for the data in all three sessions, then the research will proceed to compare signal data from different subjects for a relatively lower correlation coefficient. Data acquired in preliminary simulations show a distinctive contrast between subjects and in this study the primary focus will be finding a strong similarity among the brain signals acquired for the same subject. So in conclusion each subject’s data will be correlated for the same task (eyes opened and eyes closed) across three separate sessions then the one subject’s data will be correlated with another and the findings will be presented in the symposium.