Movement analysis is a scientific discipline with the aim of collecting information on the biomechanics of the musculoskeletal system during a motor act. In the clinical and sport fields, this discipline can be applied to commonly performed acts, such as running and walking, in order to highlight the presence of motor deficits in an individual. This can be particularly useful for people with specific kinds of frailties, both motor and intellectual. In fact, people with intellectual disabilities can find certain common tasks more difficult and less immediate than non-disabled people do, and movement analysis can be effectively used to correct any execution errors that could lead to injuries. Nevertheless, studies on these subjects are few, due especially to the difficulty of these individuals in understanding the assigned tasks. This study started with the development of an algorithm in the Matlab environment, with the aim of extracting the main parameters of running and jumping from inertial sensors worn on the waist through dedicated belts. Then, the developed algorithm was applied to data collected on a group of individuals suffering from various types of intellectual disabilities. The study subjects were selected thanks to the collaboration of the PlayMore sports centre, where they practice running weekly. The tests were performed on their usual practice day, both before training and afterwards, in order to take into account any possible differences due to fatigue. Then, tests were performed with the same procedure after a month, in order to check if a constant practice could lead to any improvement. In order to collect data, two different inertial sensors were used: the OPAL sensor and the G-WALK sensor. To assess if the data extracted from the two kind of sensors were comparable, validation tests were performed by fixing both sensors to a same rigid object; the object was initially moved in an oscillatory motion and then held still in the hands of a walking subject. This comparison leads to say that, using the same algorithm on the extracted data, the two sensors can be used indifferently. In fact, excluding the first performed test, which was inconclusive due to the incorrect positioning of the sensors, very similar trends and low standard deviation values were obtained using the second test. Afterwards, tests were performed following a specific protocol: the subjects, after performing a static jump and a counter-movement jump, had to run along a 130-meter-long predetermined route. Subsequently, the acceleration signals in all directions were processed, both for running and jumping. The algorithm developed to analyse the running signal allowed as a first thing to detect the initial contact and last contact instants of the foot on the ground. The initial contact instants were identified as the antero-posterior acceleration’s positive peaks located just before a negative peak, while the last contact instants were identified as the instants when the descending vertical acceleration reaches the value of gravitational acceleration. The detection of these events allowed to derive the spatiotemporal parameters of interest, namely the step time, the swing time, and the stance time. The distributions obtained for these parameters were then compared among the different running situations to study their trend over time. Then, other parameters of interest were computed, as listed below: - The percentage of the route performed while running. - The asymmetry index, computed for both the step time and the stance time. - The amplitude variability of the acceleration, the velocity, and the displacement. These parameters provide the variability among the individual steps of a same subject. The velocity and displacement signals were obtained from subsequent integrations of the acceleration signal with manual correction of the drift. - The Harmonic Ratio. This parameter was obtained from the Fourier transform and it was computed for all three directions. - The Approximate Entropy. The last two computed parameters, according to literature, provide an estimate of the symmetry and regularity of a signal. Since they were never previously applied to running gait, but only to walking gait, they needed validation. Tests were thus performed by applying increasingly heavier weights to the right ankle, in order to control the variation of the symmetry of the signal; the trend of the two parameters was then studied in order to assess whether or not they could be used in this study. By carrying out these tests, the procedure for the calculation of the harmonic ratio for running gait was validated as a measure of the symmetry of the step; contrariwise for the approximate entropy. In fact, the statistical tests performed (significance threshold p=0.05) showed significant differences in all three directions for the distributions of the harmonic ratio values, highlighting a significant decrease in the value of this parameter as the applied load increases, which corresponds to an increase in the asymmetry of the step. On the contrary, the approximate entropy distributions had a less regular trend, in fact the differences between the three load situations are not always significant. Moreover, in one case, while comparing the unloaded situation and the situation with a one-kilogram weight applied for the medio-lateral direction, there is a significant difference in the opposite direction of what expected: the values decrease as the load increases. Therefore, only the harmonic ratio was considered as an indicator of the symmetry of the running gait for the tests of this study, and the approximate entropy was ignored. For the jump analysis, the beginning and the end of each jump were identified on the vertical acceleration signal. In order to do so, the positive peaks respectively before and after the swing phase were considered. The swing phase is characterized, in the signal, by a vertical acceleration approximately constant and equal to the value of gravitational acceleration. From the thus identified instants, the length of each jump was obtained. The jump lengths were then compared among the different situations, taking into account the fact that, in a high jump, the reached height is proportional to the time length squared, following the relationship Δy=1/2 g∙t^2, where Δy is the elevation, t is the time necessary to reach the highest point of the trajectory (which is equal to half of the flight time) and g is the gravitational acceleration. For the running tests, an intersubjective statistical analysis was carried out on all the subjects of the study to check if there were any recognizable trends in the analysed parameters which could be explained in light of the literature search. However, the results obtained were not always significant and justifiable, therefore an intrasubjective statistical analysis was performed to highlight any unusual behaviour in a single subject and try to give an explanation. From the statistical tests performed it was found that, for most of the subjects and after both the considered training sessions, the stance time increases and, consequently, the speed decreases, probably due to accumulated fatigue. In addition, a decrease in fatigue was highlighted after one month of exercise; in fact, in most of the cases, the slowdown was less marked during the second test performed. This improvement in fatigue resistance, after one month of training, is confirmed by the intersubjective analysis. The reason why some subjects’ velocity was higher after the training session can be explained by considering the resting time that, for them, elapsed between the end of the training session and the start of the running test, due to the limited number of available sensors. Another hypothesis could be that this acceleration was caused, for some subjects, by the alternation of walking and running. The trend of the amplitude variability of the acceleration was, in most cases, in line with the trend of the subject’s speed: the faster is the running, the greater is the variability of the acceleration. On the contrary, the amplitude variability of speed and displacement showed more random trends, probably due to integration errors. The vertical oscillation parameter also didn’t show a trend in line with the speed variation, always due to possible errors in the double integration performed. It was found that, in most of the cases, the harmonic ratio varied consistently with the asymmetry index, in accordance with the assumption made of using this parameter as an index of the symmetry of the running gait. Studying the trend of this parameter, it was found that after one month there was no improvement in the symmetry of the running gait. A statistical analysis was carried out also for the jump tests, but no significant difference was found between the duration of the analysed jumps. This lack of significance may be due to the difficulty in understanding, for some subjects, the difference between the two jumps considered. Other problems could be the low number of jumps analysed and an actual lack of positive effects on the jump’s duration and elevation after only one month of running training. To improve the study and to obtain more reliable results, the number of subjects examined should be increased and a group with more homogeneous disabilities should be considered. In addition, the tests should be carried out immediately after the end of the training session, in such a way as to not allow the subject to rest. In order to do this, a greater number of sensors would be needed, or the start and the end of the subjects training sessions should be staggered. Other tips to reduce experimental errors could be: tie the sensor belts tighter directly on the waist of the subjects and not over several layers of clothing; the subjects should run individually so as not to influence the speed of the running gait; more time should be spent to explain how to perform the two different kinds of jump and the comprehension should be verified with some preliminary jump tests. Furthermore, extending the monitoring time would allow to better highlight any significant changes in the subjects’ motor ability, to carry out further tests and to obtain more data on each subject. The tests of this study were carried out during the health emergency caused by the COVID-19 pandemic. For this reason, not all subjects were able to train continuously during the month under examination and this could have led to problems in the analysis of the results.
L’analisi del movimento è una disciplina scientifica finalizzata alla raccolta di informazioni sulla biomeccanica del sistema muscolo-scheletrico durante un atto motorio. Questa disciplina, in ambito clinico e sportivo, può essere applicata ad atti eseguiti comunemente, come la corsa e il cammino, al fine di evidenziare la presenza di deficit motori in un individuo. Ciò può essere particolarmente utile in presenza di individui con fragilità specifiche, non solo motorie ma anche intellettive. Per i soggetti con disabilità intellettive, infatti, determinati atti motori possono risultare più difficili e meno immediati rispetto a quanto avviene per i normodotati e l’analisi del movimento può essere efficacemente impiegata per correggere eventuali difetti d’esecuzione che potrebbero portare ad infortuni. Ciononostante, gli studi su questi soggetti sono ancora in numero estremamente limitato a causa, in particolare, della difficoltà degli individui nel comprendere i compiti assegnati. Il presente studio si è in primo luogo occupato di sviluppare un algoritmo in ambiente Matlab per poter estrarre parametri relativi alla corsa e al salto da sensori inerziali indossati in vita tramite apposite cinture. In seguito, l’algoritmo sviluppato è stato applicato ai dati raccolti analizzando la corsa e il salto di un gruppo di ragazzi affetti da varie tipologie di disabilità intellettive. I ragazzi sono stati selezionati grazie alla collaborazione del centro sportivo PlayMore, presso cui si allenano una volta a settimana nella corsa. Le prove con i sensori sono state eseguite il giorno del loro allenamento, sia prima che dopo la seduta di allenamento per tener conto di eventuali differenze dovute alla stanchezza. In seguito, le prove sono state effettuate nuovamente dopo un mese con le stesse modalità, per verificare eventuali miglioramenti dovuti all’allenamento costante sostenuto. Per raccogliere i dati sono state utilizzate due diverse tipologie di sensori inerziali: il sensore OPAL e il sensore G-WALK. Per verificare se i dati estratti dai due diversi sensori fossero confrontabili sono state effettuate delle prove di validazione legando entrambi i sensori ad uno stesso oggetto rigido, prima in movimento oscillatorio in mano ad un soggetto fisso, poi fermo in mano al soggetto in movimento. Da questo confronto si è concluso che, utilizzando lo stesso algoritmo sui dati ricavati dai due diversi sensori, è possibile utilizzare i due sensori indifferentemente. Infatti, escludendo la prima prova di confronto effettuata, inconclusiva a causa del posizionamento errato dei sensori, con la seconda prova si sono ottenuti andamenti molto simili e valori di scarto quadratico medio bassi. In seguito, le prove si sono svolte seguendo un preciso protocollo: i soggetti, dopo aver effettuato un salto in alto statico e un salto in alto con contromovimento, hanno percorso un tragitto predefinito lungo 130 m. In seguito, è stato elaborato il segnale relativo all’accelerazione nelle tre direzioni sia per la corsa che per il salto. Per la corsa, l’algoritmo sviluppato ha consentito in primo luogo l’identificazione degli istanti di contatto iniziale e finale del piede al terreno. Gli istanti di contatto iniziale sono stati identificati come i picchi positivi dell’accelerazione antero-posteriore immediatamente precedenti ad un picco negativo, mentre gli istanti di contatto finale sono stati identificati come il punto in cui l’accelerazione verticale discendente raggiunge il valore dell’accelerazione gravitazionale. L’identificazione di questi eventi ha consentito di ricavare i parametri spazio-temporali d’interesse, cioè il tempo del passo semplice, il tempo di volo e il tempo di appoggio. Le distribuzioni ottenute per questi parametri sono state confrontate fra le diverse situazioni di corsa per studiarne l’andamento nel tempo. Sono in seguito stati calcolati gli altri parametri di interesse, elencati di seguito: - La percentuale di tragitto effettuata in corsa. - L’indice di asimmetria, calcolato sia per il tempo del passo semplice che per il tempo di appoggio. - La variabilità in ampiezza dell’accelerazione, della velocità e dello spostamento. Questa variabilità rappresenta la variabilità presente tra i singoli passi di un soggetto. I grafici di velocità e spostamento sono stati ottenuti a partire dal grafico dell’accelerazione per integrazioni successive con correzione manuale della deriva. - Il rapporto armonico. Questo parametro è stato ricavato dalla trasformata di Fourier del segnale ed è stato calcolato per tutte e tre le direzioni. - L’entropia approssimativa. Gli ultimi due parametri calcolati, da letteratura, forniscono una stima rispettivamente della simmetria e della regolarità di un segnale. Poiché non sono mai stati utilizzati in precedenza in uno studio della corsa ma esclusivamente per il cammino, è stata svolta una validazione. Sono state quindi effettuate delle prove di corsa su un soggetto normodotato applicando un peso crescente alla caviglia destra, in modo tale da controllare la variazione della simmetria del segnale, e si è studiato l’andamento dei due parametri per deciderne l’utilizzabilità nello studio. Effettuando tali prove, è stato possibile validare la procedura di calcolo del rapporto armonico per la corsa come una misura della simmetria del passo, ma non quella dell’entropia approssimativa. Infatti, i test statistici condotti (soglia di significatività p=0.05) hanno evidenziato differenze significative in tutte e tre le direzioni per le distribuzioni dei valori del rapporto armonico, riscontrando una diminuzione significativa del valore di questo parametro all’aumentare del peso applicato, che corrisponde ad un aumento dell’asimmetria del passo. Un andamento meno regolare è stato visto, invece, per le distribuzioni dell’entropia approssimativa, in cui le differenze tra le tre situazioni di carico non sono sempre significative e in un caso, in direzione medio-laterale tra la situazione senza carico e quella con il peso di un chilogrammo applicato, è presente una differenza significativa in verso opposto a quello atteso: i valori diminuiscono all’aumentare del peso. Si è deciso quindi di considerare, come indicatore della simmetria della corsa per le prove di questo studio, solo il parametro del rapporto armonico e di escludere quello dell’entropia approssimativa. Per quanto riguarda l’analisi dei salti, studiando il segnale di accelerazione in direzione verticale sono stati identificati l’inizio e la fine di ciascun salto. Per fare ciò, sono stati considerati i picchi positivi posti rispettivamente prima e dopo la fase di volo, caratterizzata nel segnale da un’accelerazione verticale circa costante e pari al valore dell’accelerazione gravitazionale. A partire dagli istanti così identificati, è stata calcolata la durata di ciascun salto. Le durate dei salti sono poi state confrontate nelle diverse situazioni, tenendo conto del fatto che, in un salto in alto, l’altezza raggiunta è proporzionale in modo quadratico alla sua durata secondo la relazione Δy=1/2 g∙t^2, in cui Δy è l’innalzamento, t è il tempo necessario a raggiungere il punto massimo della traiettoria (pari alla metà del tempo di volo) e g è l’accelerazione gravitazionale. Per quanto riguarda le prove di corsa, è stata effettuata un’analisi statistica intersoggettiva su tutti i soggetti dello studio per verificare se ci fossero andamenti riconoscibili nei parametri analizzati, spiegabili con quanto noto dalla teoria. Tuttavia, visti i risultati non sempre significativi e giustificabili ottenuti, si è deciso di procedere anche ad un’analisi statistica intrasoggettiva, per evidenziare eventuali comportamenti anomali nel singolo soggetto e provare a darne una spiegazione. Dai test statistici effettuati è stato riscontrato, per la maggior parte dei soggetti e dopo entrambi gli allenamenti considerati, un aumento del tempo di appoggio e, di conseguenza, una diminuzione della velocità, probabilmente a causa della stanchezza accumulata. Inoltre, è stata evidenziata una diminuzione nell’affaticamento dopo un mese di esercizio; infatti, nella maggior parte dei casi, il rallentamento è stato meno marcato durante la seconda prova analizzata. Tale miglioramento nella resistenza alla fatica, dopo un mese di allenamento, è confermato dall’analisi intersoggettiva svolta. Il fatto che la velocità di alcuni soggetti fosse maggiore alla fine della sessione di prova è stato spiegato considerando il tempo di riposo che, per loro, è intercorso tra la fine dell’allenamento e l’inizio della prova di corsa. Infatti, avendo a disposizione un numero limitato di sensori, alcuni soggetti dovevano aspettare che altri terminassero la prova. Un’altra ipotesi è che questa accelerazione fosse causata dall’alternarsi, per alcuni soggetti, di tratti in cammino a tratti in corsa. L’andamento della variabilità dell’accelerazione è stato riscontrato, nella maggior parte dei casi, in linea con quello della velocità del soggetto: più l’andatura della corsa è veloce, maggiore è la variabilità dell’accelerazione. Per quanto riguarda la variabilità della velocità e dello spostamento, è stata riscontrata una maggiore casualità negli andamenti, dovuta probabilmente ad errori di integrazione. Anche per il parametro dell’oscillazione verticale non è stato trovato un andamento che rispecchiasse quello della velocità, sempre a causa di possibili errori nella doppia integrazione effettuata. In accordo con la premessa fatta di utilizzare il rapporto armonico come parametro che indicasse la simmetria della corsa, è stato notato come, nella maggior parte dei casi, questo parametro variava coerentemente con l’indice di asimmetria. Studiando l’andamento del parametro, a distanza di un mese, non si è riscontrato un miglioramento nella simmetria della corsa. Anche per le prove di salto è stata effettuata un’analisi statistica, non trovando nessuna differenza significativa tra la durata dei salti analizzati all’inizio e alla fine del periodo di allenamento considerato. Tale assenza di significatività può essere dovuta alla difficoltà nel far comprendere ad alcuni dei soggetti le diverse modalità di salto (e quindi alla variabilità dell’esecuzione), al basso numero di salti analizzati, e ad un’effettiva mancanza di effetti positivi sulla durata e sull’altezza dei salti di un solo mese di allenamento nella corsa. Per migliorare lo studio e ottenere risultati più attendibili e completi si dovrebbe aumentare la numerosità dei soggetti presi in esame e considerare un gruppo più omogeneo dal punto di vista della disabilità. Inoltre, si dovrebbe fare in modo di effettuare le prove di questo studio subito dopo la fine degli allenamenti considerati, per non lasciare il tempo ai soggetti di riposare. Per fare ciò servirebbe un numero di sensori maggiore oppure scaglionare la partenza e l’arrivo dei soggetti. Altri accorgimenti da prendere per ridurre alcuni errori sperimentali sono: legare le cinture dei sensori a contatto con la vita dei soggetti e non sopra diversi strati di vestiti; far correre i soggetti individualmente per non influenzare la velocità della corsa; dedicare più tempo alle spiegazioni su come effettuare le due modalità di salto e verificare l’apprendimento con dei salti di prova. Inoltre, allungare il tempo di monitoraggio permetterebbe di evidenziare meglio eventuali cambiamenti significativi nell’abilità motoria dei soggetti, di svolgere prove ulteriori e di ottenere un numero maggiore di dati su ciascun soggetto. Le prove di questo studio sono state svolte durante l’emergenza sanitaria per il coronavirus SARS-CoV-2. Si sottolinea che non tutti i soggetti sono riusciti ad allenarsi con continuità nel corso del mese preso in esame e questo potrebbe aver portato a problematiche nell’analisi dei risultati.
Monitoraggio della corsa e del salto in soggetti con disabilità intellettiva mediante sensori inerziali indossabili
Grassi, Cecilia;Colombo, Maria
2020/2021
Abstract
Movement analysis is a scientific discipline with the aim of collecting information on the biomechanics of the musculoskeletal system during a motor act. In the clinical and sport fields, this discipline can be applied to commonly performed acts, such as running and walking, in order to highlight the presence of motor deficits in an individual. This can be particularly useful for people with specific kinds of frailties, both motor and intellectual. In fact, people with intellectual disabilities can find certain common tasks more difficult and less immediate than non-disabled people do, and movement analysis can be effectively used to correct any execution errors that could lead to injuries. Nevertheless, studies on these subjects are few, due especially to the difficulty of these individuals in understanding the assigned tasks. This study started with the development of an algorithm in the Matlab environment, with the aim of extracting the main parameters of running and jumping from inertial sensors worn on the waist through dedicated belts. Then, the developed algorithm was applied to data collected on a group of individuals suffering from various types of intellectual disabilities. The study subjects were selected thanks to the collaboration of the PlayMore sports centre, where they practice running weekly. The tests were performed on their usual practice day, both before training and afterwards, in order to take into account any possible differences due to fatigue. Then, tests were performed with the same procedure after a month, in order to check if a constant practice could lead to any improvement. In order to collect data, two different inertial sensors were used: the OPAL sensor and the G-WALK sensor. To assess if the data extracted from the two kind of sensors were comparable, validation tests were performed by fixing both sensors to a same rigid object; the object was initially moved in an oscillatory motion and then held still in the hands of a walking subject. This comparison leads to say that, using the same algorithm on the extracted data, the two sensors can be used indifferently. In fact, excluding the first performed test, which was inconclusive due to the incorrect positioning of the sensors, very similar trends and low standard deviation values were obtained using the second test. Afterwards, tests were performed following a specific protocol: the subjects, after performing a static jump and a counter-movement jump, had to run along a 130-meter-long predetermined route. Subsequently, the acceleration signals in all directions were processed, both for running and jumping. The algorithm developed to analyse the running signal allowed as a first thing to detect the initial contact and last contact instants of the foot on the ground. The initial contact instants were identified as the antero-posterior acceleration’s positive peaks located just before a negative peak, while the last contact instants were identified as the instants when the descending vertical acceleration reaches the value of gravitational acceleration. The detection of these events allowed to derive the spatiotemporal parameters of interest, namely the step time, the swing time, and the stance time. The distributions obtained for these parameters were then compared among the different running situations to study their trend over time. Then, other parameters of interest were computed, as listed below: - The percentage of the route performed while running. - The asymmetry index, computed for both the step time and the stance time. - The amplitude variability of the acceleration, the velocity, and the displacement. These parameters provide the variability among the individual steps of a same subject. The velocity and displacement signals were obtained from subsequent integrations of the acceleration signal with manual correction of the drift. - The Harmonic Ratio. This parameter was obtained from the Fourier transform and it was computed for all three directions. - The Approximate Entropy. The last two computed parameters, according to literature, provide an estimate of the symmetry and regularity of a signal. Since they were never previously applied to running gait, but only to walking gait, they needed validation. Tests were thus performed by applying increasingly heavier weights to the right ankle, in order to control the variation of the symmetry of the signal; the trend of the two parameters was then studied in order to assess whether or not they could be used in this study. By carrying out these tests, the procedure for the calculation of the harmonic ratio for running gait was validated as a measure of the symmetry of the step; contrariwise for the approximate entropy. In fact, the statistical tests performed (significance threshold p=0.05) showed significant differences in all three directions for the distributions of the harmonic ratio values, highlighting a significant decrease in the value of this parameter as the applied load increases, which corresponds to an increase in the asymmetry of the step. On the contrary, the approximate entropy distributions had a less regular trend, in fact the differences between the three load situations are not always significant. Moreover, in one case, while comparing the unloaded situation and the situation with a one-kilogram weight applied for the medio-lateral direction, there is a significant difference in the opposite direction of what expected: the values decrease as the load increases. Therefore, only the harmonic ratio was considered as an indicator of the symmetry of the running gait for the tests of this study, and the approximate entropy was ignored. For the jump analysis, the beginning and the end of each jump were identified on the vertical acceleration signal. In order to do so, the positive peaks respectively before and after the swing phase were considered. The swing phase is characterized, in the signal, by a vertical acceleration approximately constant and equal to the value of gravitational acceleration. From the thus identified instants, the length of each jump was obtained. The jump lengths were then compared among the different situations, taking into account the fact that, in a high jump, the reached height is proportional to the time length squared, following the relationship Δy=1/2 g∙t^2, where Δy is the elevation, t is the time necessary to reach the highest point of the trajectory (which is equal to half of the flight time) and g is the gravitational acceleration. For the running tests, an intersubjective statistical analysis was carried out on all the subjects of the study to check if there were any recognizable trends in the analysed parameters which could be explained in light of the literature search. However, the results obtained were not always significant and justifiable, therefore an intrasubjective statistical analysis was performed to highlight any unusual behaviour in a single subject and try to give an explanation. From the statistical tests performed it was found that, for most of the subjects and after both the considered training sessions, the stance time increases and, consequently, the speed decreases, probably due to accumulated fatigue. In addition, a decrease in fatigue was highlighted after one month of exercise; in fact, in most of the cases, the slowdown was less marked during the second test performed. This improvement in fatigue resistance, after one month of training, is confirmed by the intersubjective analysis. The reason why some subjects’ velocity was higher after the training session can be explained by considering the resting time that, for them, elapsed between the end of the training session and the start of the running test, due to the limited number of available sensors. Another hypothesis could be that this acceleration was caused, for some subjects, by the alternation of walking and running. The trend of the amplitude variability of the acceleration was, in most cases, in line with the trend of the subject’s speed: the faster is the running, the greater is the variability of the acceleration. On the contrary, the amplitude variability of speed and displacement showed more random trends, probably due to integration errors. The vertical oscillation parameter also didn’t show a trend in line with the speed variation, always due to possible errors in the double integration performed. It was found that, in most of the cases, the harmonic ratio varied consistently with the asymmetry index, in accordance with the assumption made of using this parameter as an index of the symmetry of the running gait. Studying the trend of this parameter, it was found that after one month there was no improvement in the symmetry of the running gait. A statistical analysis was carried out also for the jump tests, but no significant difference was found between the duration of the analysed jumps. This lack of significance may be due to the difficulty in understanding, for some subjects, the difference between the two jumps considered. Other problems could be the low number of jumps analysed and an actual lack of positive effects on the jump’s duration and elevation after only one month of running training. To improve the study and to obtain more reliable results, the number of subjects examined should be increased and a group with more homogeneous disabilities should be considered. In addition, the tests should be carried out immediately after the end of the training session, in such a way as to not allow the subject to rest. In order to do this, a greater number of sensors would be needed, or the start and the end of the subjects training sessions should be staggered. Other tips to reduce experimental errors could be: tie the sensor belts tighter directly on the waist of the subjects and not over several layers of clothing; the subjects should run individually so as not to influence the speed of the running gait; more time should be spent to explain how to perform the two different kinds of jump and the comprehension should be verified with some preliminary jump tests. Furthermore, extending the monitoring time would allow to better highlight any significant changes in the subjects’ motor ability, to carry out further tests and to obtain more data on each subject. The tests of this study were carried out during the health emergency caused by the COVID-19 pandemic. For this reason, not all subjects were able to train continuously during the month under examination and this could have led to problems in the analysis of the results.| File | Dimensione | Formato | |
|---|---|---|---|
|
Tesi-Colombo-Grassi.pdf
Open Access dal 01/04/2022
Dimensione
6.4 MB
Formato
Adobe PDF
|
6.4 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/175392