Muscle fatigue is a physiological condition characterized by a progressive reduction in the ability of a muscle to generate strength and plays a fundamental role in rehabilitation. Its evaluation allows not only to monitor the functional state of the muscle over time, but also to prevent overloads that could compromise recovery and increase the risk of relapses. There are several methods to assess muscle fatigue. Some of the most commonly used include analysis of metabolite concentration, assessment of maximum voluntary contraction, subjective perception of effort and duration of resistance. Among these, the analysis of surface electromyographic signals (s–EMG), acquired by non–invasive surface electrodes applied to the skin, is a technique that allows to obtain information related not only to muscle fatigue, but also motor recruitment and neuromuscular control. However, the non–stationary stochastic nature of signals requires the use of analytical tools that can observe simultaneous variations in time and frequency components to accurately describe the state of muscle fatigue. In this context, the discrete wavelet transform (DWT) has proven to be more effective than the traditional Fourier transform in detecting spectral variations associated with fatigue. In this study, carried out in collaboration with the Laboratory for Analysis of Movement at the Rota Centre in Bergamo, the electromyographic tracts relating to the medial vastus, lateral vastus and rectus femoris muscles of both the healthy limb and the injured limb were acquired in subjects undergoing anterior cruciate ligament reconstruction. The raw signals, collected at the beginning of the project (T0), after three weeks (T1) and after six weeks (T2), were pre–processed and normalized to the maximum recorded value. A 5–second time window centered on the peak of the signal was then extracted, with the aim of isolating the phase of maximum muscle activation. Two distinct indices were calculated from the signals obtained: amplitude, in the time domain and through the root mean square (RMS), and energy, in the time–frequency domain and through the DWT with the decomposition of the signal into frequency bands. The obtained values were finally statistically analysed through the Friedman test and the Wilcoxon signed–rank test, applied with two different levels of significance according to the number of subjects available in the three experimental sessions. The results show that the DWT–based approach is effective in detecting muscle fatigue during the rehabilitation protocol. However, a number of statistically significant differences were found between the healthy limb and the injured limb at time T2, suggesting that the time interval considered is not sufficient to show a functional recovery of the injured limb and emphasizing the need to extend the duration of the rehabilitation path.
La fatica muscolare è una condizione fisiologica caratterizzata da una progressiva riduzione della capacità di un muscolo di generare forza e riveste un ruolo fondamentale in ambito riabilitativo. La sua valutazione consente non solo di monitorare nel tempo lo stato funzionale del muscolo, ma anche di prevenire sovraccarichi che potrebbero compromettere il recupero e aumentare il rischio di recidive. Esistono diverse metodologie che permettono di valutare l’affaticamento muscolare. Alcuni dei metodi più comunemente utilizzati includono l’analisi della concentrazione di metaboliti, la valutazione della massima contrazione volontaria, la percezione soggettiva dello sforzo e la durata del tempo di resistenza. Tra queste, l’analisi dei segnali elettromiografici superficiali (s–EMG), acquisiti mediante elettrodi superficiali non invasivi applicati sulla cute, rappresenta una tecnica che permette di ottenere informazioni relative non solo alla fatica muscolare, ma anche al reclutamento motorio e al controllo neuromuscolare. Tuttavia, la natura stocastica non stazionaria dei segnali impone l’impiego di strumenti di analisi in grado di osservare variazioni simultanee nelle componenti temporali e frequenziali, al fine di descrivere accuratamente lo stato di affaticamento muscolare. In questo contesto, la trasformata wavelet discreta (DWT) si è dimostrata più efficace rispetto alla tradizionale trasformata di Fourier nell’individuare variazioni spettrali associate alla fatica. Nel presente studio, realizzato in collaborazione con il Laboratorio di Analisi del Movimento del Centro Rota di Bergamo, sono stati acquisiti i tracciati elettromiografici relativi ai muscoli vasto mediale, vasto laterale e retto femorale, sia dell’arto sano che di quello infortunato, in soggetti sottoposti a ricostruzione del legamento crociato anteriore. I segnali grezzi, raccolti all’inizio del progetto (T0), dopo tre settimane (T1) e dopo sei settimane (T2), sono stati pre–elaborati e normalizzati rispetto al valore massimo registrato. È stata poi estratta una finestra temporale di 5 secondi centrata sul picco massimo del segnale, con l’obiettivo di isolare la fase di massima attivazione muscolare. Dai segnali così ottenuti sono stati calcolati due distinti indici: l’ampiezza, nel dominio del tempo e mediante il root mean square (RMS), e l’energia, nel dominio tempo–frequenza e attraverso la DWT con la decomposizione del segnale in bande di frequenza. I valori ottenuti sono stati infine analizzati statisticamente mediante il test di Friedman ed il test dei ranghi con segno di Wilcoxon, applicati con due differenti livelli di significatività in funzione del numero di soggetti disponibili nelle tre sessioni sperimentali. I risultati evidenziano come l’approccio basato sulla DWT sia efficace nell’individuare la presenza di affaticamento muscolare nel corso del protocollo riabilitativo. Tuttavia, sono state rilevate numerose differenze statisticamente significative tra l’arto sano e quello infortunato al tempo T2, suggerendo che l’intervallo temporale considerato non sia sufficiente ad evidenziare un recupero funzionale dell’arto leso e sottolineando la necessità di prolungare la durata del percorso riabilitativo.
Analisi tempo-frequenza di segnali EMG per lo studio della fatica muscolare in ambito riabilitativo
LOCARINI, GIORGIO
2024/2025
Abstract
Muscle fatigue is a physiological condition characterized by a progressive reduction in the ability of a muscle to generate strength and plays a fundamental role in rehabilitation. Its evaluation allows not only to monitor the functional state of the muscle over time, but also to prevent overloads that could compromise recovery and increase the risk of relapses. There are several methods to assess muscle fatigue. Some of the most commonly used include analysis of metabolite concentration, assessment of maximum voluntary contraction, subjective perception of effort and duration of resistance. Among these, the analysis of surface electromyographic signals (s–EMG), acquired by non–invasive surface electrodes applied to the skin, is a technique that allows to obtain information related not only to muscle fatigue, but also motor recruitment and neuromuscular control. However, the non–stationary stochastic nature of signals requires the use of analytical tools that can observe simultaneous variations in time and frequency components to accurately describe the state of muscle fatigue. In this context, the discrete wavelet transform (DWT) has proven to be more effective than the traditional Fourier transform in detecting spectral variations associated with fatigue. In this study, carried out in collaboration with the Laboratory for Analysis of Movement at the Rota Centre in Bergamo, the electromyographic tracts relating to the medial vastus, lateral vastus and rectus femoris muscles of both the healthy limb and the injured limb were acquired in subjects undergoing anterior cruciate ligament reconstruction. The raw signals, collected at the beginning of the project (T0), after three weeks (T1) and after six weeks (T2), were pre–processed and normalized to the maximum recorded value. A 5–second time window centered on the peak of the signal was then extracted, with the aim of isolating the phase of maximum muscle activation. Two distinct indices were calculated from the signals obtained: amplitude, in the time domain and through the root mean square (RMS), and energy, in the time–frequency domain and through the DWT with the decomposition of the signal into frequency bands. The obtained values were finally statistically analysed through the Friedman test and the Wilcoxon signed–rank test, applied with two different levels of significance according to the number of subjects available in the three experimental sessions. The results show that the DWT–based approach is effective in detecting muscle fatigue during the rehabilitation protocol. However, a number of statistically significant differences were found between the healthy limb and the injured limb at time T2, suggesting that the time interval considered is not sufficient to show a functional recovery of the injured limb and emphasizing the need to extend the duration of the rehabilitation path.| File | Dimensione | Formato | |
|---|---|---|---|
|
Locarini_Giorgio.pdf
accesso aperto
Dimensione
3.55 MB
Formato
Adobe PDF
|
3.55 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/87274