Olcay, B. OrkanPehlivan, Murat2024-08-252024-08-252023979-8-3503-4355-72165-0608https://doi.org/10.1109/SIU59756.2023.10223913https://hdl.handle.net/11454/10031031st IEEE Conference on Signal Processing and Communications Applications (SIU) -- JUL 05-08, 2023 -- Istanbul Tech Univ, Ayazaga Campus, Istanbul, TURKEYOlfactory loss is commonly recognized as one of the first symptoms of Alzheimer's disease, which makes it a potential biomarker for early diagnosis. However, psychophysical and electrophysiological tests, routinely used for diagnosis, are often not sufficient for detection of olfactory loss. In this study, chemosensory evoked electroencephalography (EEG) signals were characterized via brain-computer interface feature extraction methods to classify Alzheimer's patients and healthy individuals. Results showed that combining EEG features and University of Pennsylvania Smell Identification Test (UPSIT) scores commonly used for diagnosis of olfactory loss, achieved a conspicuous success rate of nearly 92% in diagnosis. The findings suggest that the proposed approach can be considered as a new strategy for early diagnosis of Alzheimer's diseasetr10.1109/SIU59756.2023.10223913info:eu-repo/semantics/closedAccessChemosensory brain responsesElectroencephalographyBrain-computer interfacesAlzheimer's DiseaseSupport vector machineConference ObjectWOS:0010625710001462-s2.0-85173480328N/AN/A