EEG-Based Classification of Olfactory Response to Pleasant Stimuli

Area of Science:

• Neuroscience
• Olfactory perception
• Brain activity

Background:

• Olfactory perception is complex, involving multiple brain regions.
• Electroencephalography (EEG) power spectra can distinguish pleasant from unpleasant stimuli.
• Limited research exists on EEG's ability to differentiate fine levels of olfactory pleasantness.

Purpose of the Study:

• To investigate if EEG can differentiate neural responses to olfactory stimuli across a spectrum of pleasantness.
• To develop discriminative features from local brain information within specific frequency bands.
• To compare the performance of different classifiers for this task.

Main Methods:

• Utilized local brain information from theta (θ), alpha (α), and gamma (γ) frequency bands.
• Devised discriminative features for a classification approach.
• Compared four common classifiers, with Support Vector Machine (SVM) showing the best performance.

Main Results:

• The Support Vector Machine (SVM) classifier achieved the highest accuracy at 75.71%.
• EEG spectral features effectively discriminated between fine levels of perceived pleasantness.
• The classification was performed within a cross-validation procedure.

Conclusions:

• Objective discrimination of olfactory pleasantness levels is possible using EEG spectral features.
• SVM-based classification demonstrates significant potential for analyzing olfactory perception.
• This research opens avenues for objective assessment of scent-related neural responses.