Classification of pleasant and unpleasant odor imagery EEG signals

Electroencephalography (EEG) is a widely used technique that allows researchers to measure neural activity that demonstrates how the human brain reacts to various environmental stimuli or imaginations. In this study, EEG was used to determine the brain's reactions during the imagination of the most pleasant and unpleasant odors among four kinds of odors, including orange, clove, thyme, and mint. The distinguishability of brain responses to these odors was tested using the Hilbert transform, Fast Walsh-Hadamard transform, band power, and spectrogram image features. The results showed that the Hilbert Transform-based features have great potential to classify the EEG signals recorded during the imagination of the most pleasant and unpleasant odors. The proposed method achieved an average classification accuracy of 87.75% for the test data with a k-nearest neighbor classifier.