Implementation and Designing of Line-Detection System Based on Electroencephalography (EEG)

Background: In Brain-based lie detection systems which have been recently introduced as substitutes for classic lie detection systems, the procedure for recognition of guilty and innocent subjects is done by inspection of brain signals which are acquired during the specific Polygraph test. With the aim of increasing the performance, this paper presents a powerful method for detection of Guilty persons in lie detection systems using brain signals.

Materials and Methods: It was an experimental study. The employed method is based on the extraction of P300 components from brain signals. In this way, the test protocol was designed based on Odd-ball method, firstly. This test was done on 16 people and their brain signals were acquired. After preprocessing, P300 amplitude was extracted for each person from brain signals, and finally Guilty and Innocent persons were classified by comparing amplitude through Bootstrapped Amplitude Difference (BAD) method.

Results: The obtained results show that the proposed method has detected correctly 7 out of 8 guilty persons and 8 out of 8 innocent persons. Also, the validated results show the promise of the proposed approach in discrimination of guilty subjects from innocent subjects by the accuracy of 93. 75%.

Conclusions: Knowing the existence of precious information in brain signals and their relation with brain's cognitive activities and also considering the performance of the proposed method, there are enough reasons to use the proposed approach for detection of guilty persons from innocent ones. Further, in comparison with previous methods, the impact of man ability to control brain signal parametersand creating incorrect feelings has been reduced through the proposed method.