Entropy Analysis of Electroencephalography for Post-Stroke Dysphagia Assessment

Affecting over 50% of stroke patients, dysphagia is still challenging to diagnose and manage due to its complex multifactorial nature and can be the result of disruptions in the coordination of cortical and subcortical neural activity as reflected in electroencephalographic (EEG) signal patterns. Sample Entropy (SampEn), a signal complexity or predictability measure, could serve as a tool to identify any abnormalities associated with dysphagia. The present study aimed to identify quantitative dysphagia biomarkers using SampEn from EEG recordings in post-stroke patients. Sample entropy was calculated in the theta, alpha, and beta bands of EEG recordings in a repetitive swallowing task performed by three groups: 22 stroke patients without dysphagia (controls), 36 stroke patients with dysphagia, and 21 healthy age-matched individuals. Post-stroke patients, both with and without dysphagia, exhibited significant differences in SampEn compared to healthy subjects in the alpha and theta bands, suggesting widespread alterations in brain dynamics. These changes likely reflect impairments in sensorimotor integration and cognitive control mechanisms essential for effective swallowing. A significant cluster was identified in the left parietal region during swallowing in the beta band, where dysphagic patients showed higher entropy compared to healthy individuals and controls. This finding suggests altered neural dynamics in a region crucial for sensorimotor integration, potentially reflecting disrupted cortical coordination associated with dysphagia. The precise quantification of these neurophysiological alterations offers a robust and objective biomarker for diagnosing neurogenic dysphagia and monitoring therapeutic interventions by means of EEG, a non-invasive and cost-efficient technique.