A feasibility study of using event-related potential as a biometrics

• Published in: 2016 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) Vol. 2016; pp. 4547 - 4550
• Main Authors: Yih-Choung Yu, Sicheng Wang, Gabel, Lisa A.
• Format: Conference Proceeding Journal Article
• Language: English
• Published: United States IEEE 01-08-2016
• Subjects: Algorithms; Authentication; Biometrics (access control); Biometry - methods; Correlation; Correlation coefficient; Electroencephalography; Electroencephalography - methods; Event-Related Potentials, P300 - physiology; Feasibility Studies; Feature extraction; Humans; Training
• ISSN: 1557-170X
• DOI: 10.1109/EMBC.2016.7591739

The use of an individual's neural response to stimuli (the event-related potential or ERP) has potential as a biometric because it is highly resistant to fraud relative to other conventional authentication systems. P300 is an ERP in human electroencephalography (EEG) that occurs in response to an oddball stimulus when an individual is actively engaged in a target detection task. Because P300 is consistently detectable from almost every subject, it is considered a potential signal for biometric applications. This paper presents a feasibility study of using topological plots of P300 as a biometric in subject authentication. The variation in latency and location of P300 response of 24 participants performing the P300Speller task were studied. Data sets from four participants were used for algorithm training; data from the other 20 participants were used as imposters for algorithm validation. The result showed that the algorithm was able to correctly identify three out of these four participants. Validation test also proved that the algorithm was able to reject 95% of the imposters for those three authenticated participants.