Analysis of eyewitness testimony using electroencephalogram signals

Area of Science:

• Neuroscience
• Cognitive Psychology
• Biomedical Engineering

Background:

• Face recognition is a complex cognitive process studied extensively through neurocognitive research.
• Recent advancements explore electroencephalogram (EEG) signals for face identification, seeking objective markers.

Purpose of the Study:

• To investigate the feasibility of using Brain Computer Interface (BCI) strategies to detect neural markers for differentiating guilty from innocent individuals.
• To identify discriminative features within EEG signals for forensic applications.

Main Methods:

• Implemented BCI strategies for feature extraction from single-trial EEG signals.
• Utilized time and frequency domain characteristics for feature extraction.
• Employed a Support Vector Machine (SVM) classifier to evaluate feature performance.
• Analyzed EEG data from a cohort of 28 participants.

Main Results:

• Successfully extracted features from EEG signals capable of discriminating between guilty and innocent individuals.
• Demonstrated the effectiveness of SVM classification using selected EEG signal features.
• Identified specific time and frequency domain characteristics as key discriminators.

Conclusions:

• BCI strategies and EEG signal analysis show promise for objective identification of guilt or innocence in face recognition contexts.
• This approach offers a potential neuroscientific tool for forensic investigations.
• Further research can refine these methods for real-world applications.