Improving the Classification of Olfactory Brain-Computer Interface Responses by Combining EEG and EBG Signals

The sense of smell, or olfaction, can enhance brain-computer interfaces (BCIs). Different scents can be assigned to specific commands to allow users to interact with technology naturally, but challenges remain. Accurate odor delivery systems and robust algorithms for detecting and interpreting brain activity patterns are necessary. We propose combining electroencephalography (EEG) and electrobulbography (EBG) to improve classification accuracy. Our pilot study shows promising results for a new olfactory brain-computer interface (BCI) modality that combines common spatial pattern (CSP) filtration applied to EEG and EBG to classify responses to six scent stimuli in a classical oddball paradigm.