Machine Learning Classification of Migraine Using fNIRS During a Postural Task
Authors:
- Emre YorgancigilEYEmre Yorgancigil
- Gülnaz YükselenGYGülnaz Yükselen
- Roksi FranciRFRoksi Franci
- Erkan AcarEAErkan Acar
- Elif Ilgaz AydinlarEAElif Ilgaz Aydinlar
- Pinar Yalinay DikmenPDPinar Yalinay Dikmen
- Ugur UygunogluUUUgur Uygunoglu
- Aksel SivaASAksel Siva
- Abdullah ArcanAAAbdullah Arcan
- Feride Irem SimsekFSFeride Irem Simsek
- Sinem Burcu ErdoganSESinem Burcu Erdogan
- Ata AkinAAAta Akin
Open Access
- 31 August 2026
- journal article
- research article
- Published by MDPI AG in Brain Sciences
- Vol. 16 (9) , 927
Abstract
Background: Migraine diagnosis and severity staging rest on clinical interviews according to the ICHD-3 criteria, with no validated objective biomarker. Functional near-infrared spectroscopy (fNIRS) is a portable and non-invasive method, but existing fNIRS migraine classification studies remain relatively small and rely on cognitive tasks. We assessed whether prefrontal hemodynamic responses with a head-down-to-knees maneuver separate migraine patients from controls, and high- from low-severity migraine. Methods: Prefrontal fNIRS, including short separation channels, was recorded during the maneuver in 50 interictal migraine patients and 52 controls screened for the absence of migraine, serious chronic conditions and hypertension. Nine hemodynamic parameters per chromophore (HbO, Hb and HbT) across thirty channels entered a hypothesis-neutral pipeline of 270 candidate pipelines (3 chromophores × 3 feature selection strategies × 30 classifiers) with no predefined region of interest. Results: Deoxyhemoglobin features selected by embedded L1 regularization with shrinkage-regularized linear discriminant analysis separated the groups with 92% balanced accuracy on the internal hold-out (92% sensitivity, 92% specificity, ROC-AUC 0.99; permutation p = 3 × 10−4), against 87% in development-set cross-validation and 88% under nested selection cross-validation, with a selection bias of +0.03. High- vs. low-severity classification (19 high, 31 low) did not exceed chance under nested validation (45%; permutation p = 0.45). Conclusions: A wide-scale pipeline achieved a robust, validated separation of migraine from screened controls, carried by a distributed venous-weighted deoxyhemoglobin signature; specificity against other headache disorders remains to be established. Attack frequency severity was not separable above chance, indicating that scalar hemodynamic descriptors are sufficient for a categorical but not a graded contrast.Keywords
This publication has 56 references indexed in Scilit:
- Beyond the headache: autonomic reflex dysfunction and heightened sensory sensitivity contribute to orthostatic intolerance in migraineClinical Autonomic Research, 2025
- Identification of Migraine Subtypes Using Functional Near‐Infrared Spectroscopy Data: A Domain‐Based Feature ExtractionJournal of Biophotonics, 2025
- Automatic detection of migraine disease from EEG signals using bidirectional long-short term memory deep learning modelSignal, Image and Video Processing, 2022
- A Biomarker for Discriminating Between Migraine With and Without Aura: Machine Learning on Functional Connectivity on Resting-State EEGsAnnals of Biomedical Engineering, 2019
- The physiological origin of task-evoked systemic artefacts in functional near infrared spectroscopyNeuroImage, 2012
- The effect of propranolol on cerebrovascular reactivity to visual stimulation in migraineJournal of the Neurological Sciences, 2011
- NIRS-SPM: Statistical parametric mapping for near-infrared spectroscopyNeuroImage, 2009
- Cerebrovascular reactivity to hypercapnia in migraine patients measured with near-infrared spectroscopyBrain Research, 2006
- Cerebrovascular dynamics in patients with migraine: Near-infrared spectroscopy studyNeuroscience Letters, 2006
- Decreased vasoreactivity to right cerebral hemisphere pressure in migraine without aura: a near-infrared spectroscopy studyClinical Neurophysiology, 2005