Clinical EEG in people with multiple sclerosis undergoing gait rehabilitation: A deep learning analysis

Muttaleb, Ammar Alaa Hamad (2026) Clinical EEG in people with multiple sclerosis undergoing gait rehabilitation: A deep learning analysis. [Laurea magistrale], Università di Bologna, Corso di Studio in Biomedical engineering [LM-DM270] - Cesena, Documento full-text non disponibile
Il full-text non è disponibile per scelta dell'autore. (Contatta l'autore)

Abstract

Gait rehabilitation can improve mobility in people with multiple sclerosis (MS), but the associated brain activity changes remain poorly understood. This thesis aimed to investigate motor-related EEG activity in people with MS compared with healthy controls, examine EEG changes following different gait rehabilitation protocols, and explore their relationship with clinical improvement. EEG recordings during an upper-limb motor task were analyzed in 8 healthy controls and 21 MS patients who underwent one of three gait rehabilitation protocols: progressively increasing low-intensity robot-assisted gait training (LowRAGT), conventional robot-assisted gait training (RAGT), or overground gait training (OGT). Changes in alpha and beta-band event-related desynchronization (ERD) were investigated before and after rehabilitation. An interpretable deep learning (DL) model (SincShallowNet), combined with explainability techniques, was employed to discriminate EEG recordings before and after rehabilitation and to relate EEG features with clinical improvement. The ERD analysis revealed differences in motor-related EEG activity between people with MS and healthy controls, with more bilateral alpha ERD and predominantly ipsilateral beta ERD in people with MS. Following rehabilitation, LowRAGT showed the greatest increase in contralateral desynchronization, RAGT showed smaller changes, and OGT generally showed reduced desynchronization. DL analysis revealed distinct EEG features across improvement categories. Highly improving patients showed greater alpha and low-beta band relevance especially over bilateral central and centro-parietal regions. Moderately improving patients showed greater beta band relevance, particularly over ipsilateral centro-parietal regions, while non-improving patients showed greater relevance at lower frequencies. Despite the small sample, these analyses offer preliminary insights into EEG changes following rehabilitation and their possible clinical relevance

Abstract
Tipologia del documento
Tesi di laurea (Laurea magistrale)
Autore della tesi
Muttaleb, Ammar Alaa Hamad
Relatore della tesi
Correlatore della tesi
Scuola
Corso di studio
Indirizzo
CURRICULUM BIOMEDICAL ENGINEERING FOR NEUROSCIENCE
Ordinamento Cds
DM270
Parole chiave
Rehabilitation,Multiple,Sclerosis,Upper,Limb,Motor, Task,Alpha,Beta,EEG,Oscillations,Event,Related, Desynchronization,Interpretable,Deep,Learning,Explainability, Techniques
Data di discussione della Tesi
19 Novembre 2026
URI

Altri metadati

Gestione del documento: Visualizza il documento

^