Lie Detection Using Brain P300 Signal: Preliminary Results


  • Kamyar Abhari Department of Electrical and Computer Engineering, University of Manitoba
  • Reza Fazel-Rezai Department of Electrical and Computer Engineering, University of Manitoba, Institute for Biodiagnostics, National Research Council (NRC)
  • Amir Meghdadi


The traditional lie detector called “polygraph” monitors several subject physiological activities during a question-answer interview to observe how they change in comparison to normal levels. However, because the subject can control his/her physiological reactions or show anxiety for other reasons than lying, the polygraph result is not reliable and acceptable legally. In this study, we will look into a method which does not directly depend on emotional reactions of the suspect. In this technique, the key is one of the significant brain signal’s components called P300. The P300, a positive peak in EEG signals, occurs about 300 msec after the subject is confronted by a sequence of stimuli which one of them is rarely presented. In this method, EEG signals are recorded while the sequence of words, pictures or sounds (oddball paradigm) is displayed to the suspect and analyzed afterward to indicate whether or not the subject has some information about the relevant stimulus. In this study, we precisely designed an interface including series of random pictures and ask subjects to press a key immediately after observing particular images. On the other hand, we have placed our lie-related picture in this sequence and observe the user’s EEG response. Using this technique, users have no control on their P300 and so it is more reliable and accurate than polygraph. In this paper, preliminary results on lie detection using the brain P300 signal have been presented. 




How to Cite

K. Abhari, R. Fazel-Rezai, and A. Meghdadi, “Lie Detection Using Brain P300 Signal: Preliminary Results”, CMBES Proc., vol. 30, no. 1, Dec. 2007.