Migraine Detection in Young Group Based on Functional Near-Infrared Spectroscopy Measurements

• Published in: IEEE journal of selected topics in quantum electronics Vol. 31; no. 4: Adv. in Neurophoton. for Non-Inv. Brain Mon.; pp. 1 - 11
• Main Authors: Chen, Wei-Ta, Li, Chia-Chen, Liu, Yao-Hong, Cheong, Pou-Leng, Wang, Yi-Min, Sun, Chia-Wei
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.07.2025; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Anxiety disorders; Arithmetic; Blood; Cognitive tasks; Depression; Electronic mail; Feature extraction; Functional near-infrared spectroscopy; Functional near-infrared spectroscopy (fNIRS); Hospitals; Infrared spectra; Infrared spectroscopy; Machine learning; Medical imaging; mental arithmetic task (MAT); Migraine; Near infrared radiation; neurofeedback; Oxygenation; prefrontal cortex (PFC); Psychological stress; Spectrum analysis; verbal fluency task; young migraine
• ISSN: 1077-260X; 1558-4542
• DOI: 10.1109/JSTQE.2025.3540761

This study investigated the neurovascular responses in young individuals with fewer complications using functional near-infrared spectroscopy (fNIRS). Thirty-two young migraines and thirty-two healthy control subjects (HC) were measured by fNIRS to observe changes in hemoglobin in the prefrontal cortex (PFC). According to the structural changes in the frontal cortex in migraine patients, two mental stress tasks and a concentration task (CT) were designed. The statistical findings showed that all three tasks revealed differences in prefrontal blood oxygenation between groups. Specifically, during the mental task-related exercises, a significant difference was identified in the left hemisphere, whereas during the CT, a notable distinction was noted in the right hemisphere. Furthermore, machine learning techniques were applied for migraine classification, receiving test accuracies of 82%, 89%, and 90% for the mental arithmetic task (MAT), the verbal fluency task (VFT), and the CT, respectively. These results demonstrate the feasibility of utilizing fNIRS with machine learning to classify migraines in young individuals.