The aim of this study is to investigate gait characteristics in persons with multiple sclerosis (pwMS) using data acquired during the six-minute walking test (6MWT) performed both overground and on a treadmill. Inertial measurement unit (IMU) data collected at the center of mass (COM) level, as well as vertical ground reaction forces (GRFs), moments along the x and y axes (Mx, My), and center of pressure coordinates (COPx and COPy) data were recorded. Initial contact events from both the accelerometer signal and GRFs were detected, allowing for the segmentation of each minute into individual steps. Linear and non-linear measures have been examined for each minute of the test to analyze spatio-temporal parameters, variability (root mean square), stability (Lyapunov exponent), symmetry (step and stride regularity, indices of asymmetry in GRFs and COP), and complexity of gait (multiscale sample entropy). Additionally, the frequency content of the signals was analyzed (Harmonic Ratio HR, Power Spectrum, Spectral Arch Length, SPARC). The study was conducted on a sample of 25 healthy individuals (HC) composed by 13 females, 12 males, average age 29,64± 7,11 years, and 27 people with MS, 19 females, 8 males, average age 36 years ± 8,959. Statistical comparisons were made to identify differences between healthy individuals and pwMS at baseline (MS TO). The second objective of this study is to identify indicators of disease progression. Follow-up (T1) assessments were conducted one year later on a group of MS patients (18 persons), and a comparison was made with the baseline measurements (MS TO vs MS T1) and with control (HC vs MS T1). Finally, to identify predictive markers, MS patients were divided into two groups based on clinical deterioration at the follow-up assessment. Parameters extracted from segments of one minute data did not differ between the two subgroups. However, many parameters extracted over each stride and averaged over all strides of one minute showed significant differences at baseline between the two subgroups of patients. Interestingly, the number of significant metrics increased from minute 1 to minute 6, suggesting a greater impact of fatigue on the group of patients with clinical deterioration. Subsequently, the predictive ability of a machine learning algorithm (Support Vector Machine, SVM) was evaluated using significant features achieving good results. This study provides valuable insights into gait characteristics in individuals with MS using data from the 6MWT. Nonetheless, the identified gait parameters and prognostic indicators can contribute to a better understanding of disease progression and guide the development of interventions aimed at improving mobility and quality of life in MS patients. However, it is important to note that the limited sample size in this study may have implications for the accuracy of the results. These findings would benefit from future evaluation on a larger sample to enhance their validity and generalizability.
L'obiettivo di questo studio è quello di analizzare le caratteristiche del cammino nelle persone affette da sclerosi multipla (MS) utilizzando dati acquisiti durante il test di camminata di sei minuti (6MWT) eseguito sia Overground che su Treadmill. Sono stati registrati i dati dell'unità di misurazione inerziale (IMU) raccolti al livello del centro di massa (COM), così come le forze di reazione verticale (GRF), i momenti lungo gli assi x e y (Mx, My) e le coordinate del centro di pressione (COPx e COPy). Sono stati rilevati gli eventi relativi al contatto iniziale del piede al suolo sia sul segnale dell'accelerometro che sulla GRF, consentendo la suddivisione dei passi per ogni minuto. Sono state calcolate misure lineari e non al fine di analizzare i parametri spazio-temporali, la variabilità (root mean square), la stabilità (Lyapunov Exponent), la simmetria (stride e step regularity, indici di asimmetria nelle GRF e nel COP) e la complessità della camminata (Multiscale Sample Entropy). Inoltre, è stata analizzata la frequenza dei segnali (Harmonic Ratio,HR, Spettro della Potenza, Lunghezza dell'Arco Spettrale, SPARC). Lo studio è stato condotto su un campione di 25 individui sani (HC), composto da 13 donne, 12 uomini, età media 29,64 ± 7,11 anni, e 27 persone con SM, 19 donne, 8 uomini, età media 36 anni ± 8,959. Sono state effettuate comparazioni statistiche per identificare le differenze tra individui sani e persone con SM al basaline (SM TO). Il secondo obiettivo di questo studio è individuare indicatori di progressione della malattia. Le valutazioni di follow-up (T1) sono state effettuate un anno dopo su un gruppo di pazienti con SM (18 persone), e sono state confrontate con le misurazioni iniziali (SM TO vs SM T1) e con il gruppo di controllo (HC vs SM T1). Infine, per identificare marcatori predittivi, i pazienti con SM sono stati divisi in due gruppi in base alla presenza di peggioramento clinico ad un anno. I parametri estratti dai segmenti di dati di un minuto non hanno mostrato differenze significative tra i due sottogruppi. Tuttavia, molti parametri estratti sui passi di ogni minuto presentano differenze significative tra i due sottogruppi di pazienti al baseline. È interessante notare che il numero di metriche significative è aumentato dal minuto 1 al minuto 6, suggerendo un maggiore impatto della fatica sul gruppo di pazienti con peggioramento clinico. Successivamente, è stata valutata la capacità predittiva di un algoritmo di apprendimento automatico (Support Vector Machine, SVM utilizzando i parametri risultati significativi, ottenendo buoni risultati. Questo studio fornisce preziose intuizioni sulle caratteristiche della camminata nelle persone con SM utilizzando dati provenienti dal 6MWT. I parametri di camminata identificati e gli indicatori prognostici possono contribuire a una migliore comprensione della progressione della malattia e guidare lo sviluppo di interventi mirati a migliorare la mobilità e la qualità della vita nei pazienti con SM. Tuttavia, è importante notare che la limitata dimensione del campione di questo studio potrebbe avere implicazioni sull'accuratezza dei risultati. Queste scoperte trarrebbero beneficio da una futura valutazione su un campione più ampio in modo da aumentar la validità e la generalizzabilità dei risultati.
Prognostic markers in persons with Multiple Sclerosis at early stage: a multifactorial analysis of overground and treadmill walking.
MARLIANI, VIRNA
2022/2023
Abstract
The aim of this study is to investigate gait characteristics in persons with multiple sclerosis (pwMS) using data acquired during the six-minute walking test (6MWT) performed both overground and on a treadmill. Inertial measurement unit (IMU) data collected at the center of mass (COM) level, as well as vertical ground reaction forces (GRFs), moments along the x and y axes (Mx, My), and center of pressure coordinates (COPx and COPy) data were recorded. Initial contact events from both the accelerometer signal and GRFs were detected, allowing for the segmentation of each minute into individual steps. Linear and non-linear measures have been examined for each minute of the test to analyze spatio-temporal parameters, variability (root mean square), stability (Lyapunov exponent), symmetry (step and stride regularity, indices of asymmetry in GRFs and COP), and complexity of gait (multiscale sample entropy). Additionally, the frequency content of the signals was analyzed (Harmonic Ratio HR, Power Spectrum, Spectral Arch Length, SPARC). The study was conducted on a sample of 25 healthy individuals (HC) composed by 13 females, 12 males, average age 29,64± 7,11 years, and 27 people with MS, 19 females, 8 males, average age 36 years ± 8,959. Statistical comparisons were made to identify differences between healthy individuals and pwMS at baseline (MS TO). The second objective of this study is to identify indicators of disease progression. Follow-up (T1) assessments were conducted one year later on a group of MS patients (18 persons), and a comparison was made with the baseline measurements (MS TO vs MS T1) and with control (HC vs MS T1). Finally, to identify predictive markers, MS patients were divided into two groups based on clinical deterioration at the follow-up assessment. Parameters extracted from segments of one minute data did not differ between the two subgroups. However, many parameters extracted over each stride and averaged over all strides of one minute showed significant differences at baseline between the two subgroups of patients. Interestingly, the number of significant metrics increased from minute 1 to minute 6, suggesting a greater impact of fatigue on the group of patients with clinical deterioration. Subsequently, the predictive ability of a machine learning algorithm (Support Vector Machine, SVM) was evaluated using significant features achieving good results. This study provides valuable insights into gait characteristics in individuals with MS using data from the 6MWT. Nonetheless, the identified gait parameters and prognostic indicators can contribute to a better understanding of disease progression and guide the development of interventions aimed at improving mobility and quality of life in MS patients. However, it is important to note that the limited sample size in this study may have implications for the accuracy of the results. These findings would benefit from future evaluation on a larger sample to enhance their validity and generalizability.| File | Dimensione | Formato | |
|---|---|---|---|
|
Marliani_Virna.pdf
accesso riservato
Dimensione
7.06 MB
Formato
Adobe PDF
|
7.06 MB | Adobe PDF |
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/20.500.12608/46926