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

Area of Science:

• Neuroscience
• Biomedical Engineering
• Sensory Systems

Background:

• Brain-computer interfaces (BCIs) offer natural interaction with technology.
• Olfaction (sense of smell) presents a unique modality for BCI enhancement.
• Current olfactory BCIs face challenges in accurate odor delivery and brain signal interpretation.

Purpose of the Study:

• To develop and evaluate a novel olfactory brain-computer interface (BCI) modality.
• To improve the classification accuracy of brain responses to olfactory stimuli.
• To explore the combined use of electroencephalography (EEG) and electrobulbography (EBG) for olfactory BCI.

Main Methods:

• Utilized a classical oddball paradigm with six distinct scent stimuli.
• Recorded brain activity using electroencephalography (EEG) and electrobulbography (EBG).
• Applied Common Spatial Pattern (CSP) filtration to EEG and EBG data for signal processing.

Main Results:

• The combined EEG and EBG approach demonstrated promising results in pilot testing.
• The proposed method showed improved classification accuracy for olfactory responses.
• Successful classification of brain activity patterns associated with different scent stimuli was achieved.

Conclusions:

• The integration of EEG and EBG offers a viable new modality for olfactory BCIs.
• This approach enhances the potential for naturalistic human-computer interaction via smell.
• Further research is warranted to optimize odor delivery and BCI algorithms.