The functional recovery of the upper limb in neurological patients, such as stroke survivors, requires rehabilitation approaches aimed at countering the adoption of compensatory motor strategies. Although the literature highlights the potential of biofeedback to drive neuroplasticity, there is a need to validate effective paradigms that are also applicable in tele-rehabilitation contexts. Within this general framework, the present thesis work, developed as part of the PRIN VISIONARY project, aims to validate two motor learning protocols for the upper limb based on real-time sEMG biofeedback. Developed in the Unity environment, the tasks involve planar flexion-extension movements reproduced on screen by a cursor moving towards virtual targets via a motion tracking system. The first protocol involves the biceps and triceps brachii; the second involves the upper trapezius and triceps. To induce and amplify the generation of compensatory motor strategies, elastic bands calibrated individually by calculating the Electromyographic Fatigue Threshold were used. The core of the system lies in the controlled alteration of the visuo-motor relationship, modulated by the activation of a target muscle and an adaptive threshold mechanism: an excessive contraction of the triceps muscle during extension generates a radial deviation of the cursor in Motor Task 1; whereas an excessive activation of the upper trapezius causes the enlargement of a virtual object and the detachment of the cursor in Motor Task 2. In both cases, the user is forced to modulate muscle activation to complete the exercise. The protocols include three distinct experimental phases (Baseline, Adaptation, and Washout) and are structured over two days: on Day 1 (without biofeedback), the effects of the mechanical load alone (elastic band) introduced in the Adaptation phase are evaluated; on Day 2 (conducted entirely in the presence of the elastic bands), the biofeedback is introduced in the Adaptation phase to study potential motor adaptation and the adoption of any new motor strategies. The results regarding muscle activation and kinematic errors, recorded on 15 healthy subjects for each motor task, show that the elastic load induces a purely transient muscle increase and accentuates certain compensatory strategies. Furthermore, the muscle activation-based biofeedback triggers a genuine motor learning process: subjects are able to progressively reduce target muscle activity, minimizing kinematic errors and maintaining this result even in the Washout phase, when the biofeedback is removed. This study allowed for the preliminary validation of the system's effectiveness, laying the groundwork for future clinical applications aimed at countering pathological compensations and promoting functional recovery in post-stroke patients, utilizing a paradigm specifically designed for home-based tele-rehabilitation.
Il recupero funzionale dell'arto superiore in pazienti neurologici, come quelli con esiti di ictus, richiede approcci riabilitativi mirati a contrastare l'adozione di strategie motorie compensatorie. Sebbene la letteratura evidenzi il potenziale del biofeedback per guidare la neuroplasticità, vi è la necessità di validare paradigmi efficaci e applicabili anche in contesti di tele-riabilitazione. In questo quadro generale, il presente lavoro di tesi, sviluppato nell’ambito del progetto PRIN VISIONARY, ha come obiettivo la validazione di due protocolli di apprendimento motorio per l'arto superiore basati su biofeedback sEMG in tempo reale. Sviluppati in ambiente Unity, i task prevedono movimenti planari di flesso-estensione riprodotti a schermo da un cursore che si sposta verso target virtuali, tramite un sistema di motion tracking. Il primo protocollo coinvolge bicipite e tricipite brachiale; il secondo, trapezio superiore e tricipite. Per indurre ed amplificare la generazione di strategie motorie di compenso sono state utilizzate bande elastiche calibrate individualmente calcolando la Soglia di Fatica Elettromiografica. Il cuore del sistema risiede nell'alterazione controllata della relazione visuo-motoria, modulata dall’attivazione di un muscolo target e da un meccanismo a soglia adattiva: un'eccessiva contrazione del muscolo tricipite in estensione genera una deviazione radiale del cursore nel Task motorio 1; mentre un’attivazione eccessiva del trapezio superiore genera l'ingrandimento di un oggetto virtuale e il distacco del cursore nel Task motorio 2. In entrambi i casi, l'utente è costretto a modulare l'attivazione muscolare per completare l'esercizio. I protocolli prevedono tre fasi sperimentali distinte (Baseline, Adaptation e Washout) e si articolano in due giornate: nel Giorno 1 (senza biofeedback) si valutano gli effetti del solo carico meccanico (banda elastica) introdotto nella fase di Adaptation; nel Giorno 2 (condotto interamente in presenza delle bande elastiche) si introduce il biofeedback nella fase di Adaptation per studiare l’eventuale adattamento motorio e l’adozione di eventuali nuove strategie motorie. I risultati relativi all’attivazione muscolare e a gli errori cinematici, rilevati su 15 soggetti sani per ogni task motorio, mostrano che il carico elastico induce un incremento muscolare puramente transitorio ed accentua alcune strategie di compenso. Inoltre, il biofeedback basato sull’attivazione muscolare innesca un reale processo di apprendimento motorio: i soggetti riescono a ridurre progressivamente l'attività del muscolo target, minimizzando gli errori cinematici e mantenendo questo risultato anche nella fase di Washout, quando il biofeedback è rimosso. Questo studio ha consentito di validare preliminarmente l'efficacia del sistema, ponendo le basi per future applicazioni in ambito clinico finalizzate a contrastare le compensazioni patologiche e favorire il recupero funzionale in pazienti post-ictus, con un paradigma progettato per la tele-riabilitazione domiciliare.
Validazione di protocolli di apprendimento motorio in presenza di biofeedback basato su EMG per l’arto superiore
CIMMINO, DEMETRA
2025/2026
Abstract
The functional recovery of the upper limb in neurological patients, such as stroke survivors, requires rehabilitation approaches aimed at countering the adoption of compensatory motor strategies. Although the literature highlights the potential of biofeedback to drive neuroplasticity, there is a need to validate effective paradigms that are also applicable in tele-rehabilitation contexts. Within this general framework, the present thesis work, developed as part of the PRIN VISIONARY project, aims to validate two motor learning protocols for the upper limb based on real-time sEMG biofeedback. Developed in the Unity environment, the tasks involve planar flexion-extension movements reproduced on screen by a cursor moving towards virtual targets via a motion tracking system. The first protocol involves the biceps and triceps brachii; the second involves the upper trapezius and triceps. To induce and amplify the generation of compensatory motor strategies, elastic bands calibrated individually by calculating the Electromyographic Fatigue Threshold were used. The core of the system lies in the controlled alteration of the visuo-motor relationship, modulated by the activation of a target muscle and an adaptive threshold mechanism: an excessive contraction of the triceps muscle during extension generates a radial deviation of the cursor in Motor Task 1; whereas an excessive activation of the upper trapezius causes the enlargement of a virtual object and the detachment of the cursor in Motor Task 2. In both cases, the user is forced to modulate muscle activation to complete the exercise. The protocols include three distinct experimental phases (Baseline, Adaptation, and Washout) and are structured over two days: on Day 1 (without biofeedback), the effects of the mechanical load alone (elastic band) introduced in the Adaptation phase are evaluated; on Day 2 (conducted entirely in the presence of the elastic bands), the biofeedback is introduced in the Adaptation phase to study potential motor adaptation and the adoption of any new motor strategies. The results regarding muscle activation and kinematic errors, recorded on 15 healthy subjects for each motor task, show that the elastic load induces a purely transient muscle increase and accentuates certain compensatory strategies. Furthermore, the muscle activation-based biofeedback triggers a genuine motor learning process: subjects are able to progressively reduce target muscle activity, minimizing kinematic errors and maintaining this result even in the Washout phase, when the biofeedback is removed. This study allowed for the preliminary validation of the system's effectiveness, laying the groundwork for future clinical applications aimed at countering pathological compensations and promoting functional recovery in post-stroke patients, utilizing a paradigm specifically designed for home-based tele-rehabilitation.| File | Dimensione | Formato | |
|---|---|---|---|
|
Tesi_Demetra_Cimmino_Unipv.pdf
accesso aperto
Descrizione: Tesi di Laurea
Dimensione
4.57 MB
Formato
Adobe PDF
|
4.57 MB | Adobe PDF | Visualizza/Apri |
È consentito all'utente scaricare e condividere i documenti disponibili a testo pieno in UNITESI UNIPV nel rispetto della licenza Creative Commons del tipo CC BY NC ND.
Per maggiori informazioni e per verifiche sull'eventuale disponibilità del file scrivere a: [email protected].
https://hdl.handle.net/20.500.14239/35723