Hydrokinesitherapy is a rehabilitation technique performed in water, as a support to traditional land-based physiotherapy, which, through the movement of the body in water, promotes muscle relaxation, greater safety and less pain in movements and spasticity, particularly for patients with neuromotor disorders. Due to the lack of data to objectively evaluate this type of therapy in complex pathological settings, in this thesis work, a non-invasive video-based system was used, without interfering with either the patients or the therapy itself, an essential requirement for patients with serious conditions. Kinematic analysis was conducted using the MediaPipe algorithm, which detects and tracks 33 body reference points frame by frame, generating a CSV file containing the coordinates of these points over time. The main objective was to analyse the kinematics of the lower limbs during walking in water, focusing on the joint angles of the hip, knee and ankle. The selected sample included 12 paediatric patients (6 females and 6 males), with an average age of 10.25 ± 3.14 years, diagnosed with neuromotor disorders of various aetiologies and with different clinical and motor characteristics, treated at the Fondazione TOG. The analysis was divided into two groups: patients walking both on land and in water (10 patients) and patients not walking on land, but able to walk in water (2 patients). For the first group, the acquisition protocol involved recording with a GoPro HERO10 on land, filming the walk from a starting point to an end point and back to film both sides of the patient. For each patient, at least 5 steps per side were recorded, with support such as walkers or physical assistance from the therapist where necessary. In the water, filming was done with the same camera inserted into a DomePort, with buoyancy aids if required. Each acquisition was repeated 3 times, at 3-month intervals. The analysis involved calculating the Range Of Motion (ROM) of each joint, averaged over the 5 steps acquired, for both the aquatic and land conditions, then comparing the differences between ROM in water and on land. With a threshold of ±6° considered negligible, an increase in ROM in water was found for all patients, except for patient 5, characterised by hypotonic musculature. Six months after the first acquisition, 91.7% positive values were recorded compared to 51.3% at time 0. The second group included 2 patients who were not ambulatory on land but were able to walk in water. Due to the complexity of their condition, it was not possible to isolate individual steps; a linear regression model was therefore adopted to approximate the trend in joint angles. The slope of the regression line, which evaluates the prevalence of flexion over extension in the same session, and the mean value, indicative of the variation over time, were considered. For patient 2, with the clinical objective of verticalisation, neither the angular coefficient nor the mean value showed significant results, except for the change in hip angle over time. For patient 3, aimed at the recreation of the step cycle in water, the angular coefficient indicated improvements within the single session for the hip and knee angles; in addition, no temporal evolution between sessions has been demonstrated. The research has some limitations, such as the use of a single camera, the variability of its position, the manual selection of the steps and the clinical subjectivity of the patients, as well as the temporal distance of the acquisitions, which could influence the results. However, in the future, objective data on kinematics in the water of patients could be used to monitor patient progress by expanding the sample of subjects involved, to support the effectiveness of hydrokinesitherapy in contexts where other motion capture systems cannot be used.
L'idrokinesiterapia è una tecnica riabilitativa svolta in acqua, come supporto alla fisioterapia tradizionale a terra, che, attraverso il movimento del corpo in acqua, favorisce rilassamento muscolare, una maggiore sicurezza e un minor dolore nei movimenti e spasticità, in particolare per pazienti con disturbi neuromotori. Data la mancanza di dati per valutare oggettivamente questa tipologia di terapia in ambito patologico complesso, in questo lavoro di tesi è stato utilizzato un sistema video-based non invasivo, senza interferire né con i pazienti né con la terapia stessa, requisito essenziale per pazienti con condizioni gravi. L’analisi cinematica è stata condotta mediante l’algoritmo MediaPipe, che rileva e traccia 33 punti di riferimento corporei fotogramma per fotogramma, generando un file CSV contenente le coordinate di questi punti nel tempo. L’obiettivo principale era analizzare la cinematica degli arti inferiori durante la deambulazione in acqua, focalizzandosi sugli angoli articolari di anca, ginocchio e caviglia. Il campione selezionato comprendeva 12 pazienti pediatrici (6 femmine e 6 maschi), con un'età media di 10.25 ± 3.14 anni, a cui sono stati diagnosticati disturbi neuromotori di varia eziologia e con caratteristiche cliniche e motorie differenti, in cura presso Fondazione TOG. L’analisi è stata suddivisa in due gruppi: pazienti deambulanti sia a terra che in acqua (10 pazienti) e pazienti non deambulanti a terra, ma in grado di camminare in acqua (2 pazienti). Per il primo gruppo, il protocollo di acquisizione prevedeva una registrazione con GoPro HERO10 a terra, filmando il cammino da un punto di partenza a un punto di arrivo e ritorno per riprendere entrambi i lati del paziente. Per ogni paziente, si sono registrati almeno 5 passi per lato, con supporti come deambulatori o assistenza fisica del terapista, ove necessario. In acqua, le riprese sono state effettuate con la stessa telecamera inserita in un DomePort, con supporti per il galleggiamento se richiesto. Ogni acquisizione è stata ripetuta 3 volte, a intervalli di 3 mesi. L’analisi ha riguardato il calcolo del Range Of Motion (ROM) di ogni articolazione, mediato sui 5 passi acquisiti, sia per la condizione acquatica sia per quella terrestre, confrontando poi le differenze tra ROM in acqua e a terra. Con una soglia di ±6° considerata trascurabile, si è riscontrato un aumento del ROM in acqua per tutti i pazienti, tranne per il paziente 5, caratterizzato da muscolatura ipotonica. A sei mesi dalla prima acquisizione, si è registrato il 91,7% di valori positivi contro il 51,3% al tempo 0. Il secondo gruppo comprendeva 2 pazienti non deambulanti a terra ma in grado di camminare in acqua. Data la complessità delle loro condizioni, non è stato possibile isolare singoli passi; si è quindi adottato un modello di regressione lineare per approssimare l’andamento degli angoli articolari. Sono stati considerati il coefficiente angolare della retta di regressione, che valuta la prevalenza della flessione rispetto all’estensione nella stessa seduta, e il valore medio, indicativo della variazione nel tempo. Per il paziente 2, con obiettivo clinico di verticalizzazione, né il coefficiente angolare né il valore medio hanno mostrato risultati significativi, ad eccezione della variazione dell'angolo all'anca nel tempo. Per il paziente 3, con l'obiettivo di ricreare del ciclo del passo in acqua, il coefficiente angolare ha indicato miglioramenti all’interno della singola seduta per l'angolo all'anca e al ginocchio; inoltre, non è stata dimostrata una evoluzione temporale tra le diverse sedute. La ricerca presenta alcuni limiti, quali l’uso di una singola telecamera, la variabilità della sua posizione, la selezione manuale dei passi e la soggettività clinica dei pazienti, oltre alla distanza temporale delle acquisizioni, che potrebbe influenzare i risultati. Tuttavia, in futuro, dati oggettivi sulla cinematica in acqua di pazienti potrebbero essere utilizzati per monitorare i progressi dei pazienti, ampliando il campione di soggetti coinvolti, al fine di supportare l’efficacia dell’idrokinesiterapia in contesti in cui non è possibile utilizzare altri sistemi di acquisizione del movimento.
Video-based kinematic assessment of lower limbs in paediatric patients with severe neuromotor disorders during hydrokinesitherapy
Palotti, Giorgia
2023/2024
Abstract
Hydrokinesitherapy is a rehabilitation technique performed in water, as a support to traditional land-based physiotherapy, which, through the movement of the body in water, promotes muscle relaxation, greater safety and less pain in movements and spasticity, particularly for patients with neuromotor disorders. Due to the lack of data to objectively evaluate this type of therapy in complex pathological settings, in this thesis work, a non-invasive video-based system was used, without interfering with either the patients or the therapy itself, an essential requirement for patients with serious conditions. Kinematic analysis was conducted using the MediaPipe algorithm, which detects and tracks 33 body reference points frame by frame, generating a CSV file containing the coordinates of these points over time. The main objective was to analyse the kinematics of the lower limbs during walking in water, focusing on the joint angles of the hip, knee and ankle. The selected sample included 12 paediatric patients (6 females and 6 males), with an average age of 10.25 ± 3.14 years, diagnosed with neuromotor disorders of various aetiologies and with different clinical and motor characteristics, treated at the Fondazione TOG. The analysis was divided into two groups: patients walking both on land and in water (10 patients) and patients not walking on land, but able to walk in water (2 patients). For the first group, the acquisition protocol involved recording with a GoPro HERO10 on land, filming the walk from a starting point to an end point and back to film both sides of the patient. For each patient, at least 5 steps per side were recorded, with support such as walkers or physical assistance from the therapist where necessary. In the water, filming was done with the same camera inserted into a DomePort, with buoyancy aids if required. Each acquisition was repeated 3 times, at 3-month intervals. The analysis involved calculating the Range Of Motion (ROM) of each joint, averaged over the 5 steps acquired, for both the aquatic and land conditions, then comparing the differences between ROM in water and on land. With a threshold of ±6° considered negligible, an increase in ROM in water was found for all patients, except for patient 5, characterised by hypotonic musculature. Six months after the first acquisition, 91.7% positive values were recorded compared to 51.3% at time 0. The second group included 2 patients who were not ambulatory on land but were able to walk in water. Due to the complexity of their condition, it was not possible to isolate individual steps; a linear regression model was therefore adopted to approximate the trend in joint angles. The slope of the regression line, which evaluates the prevalence of flexion over extension in the same session, and the mean value, indicative of the variation over time, were considered. For patient 2, with the clinical objective of verticalisation, neither the angular coefficient nor the mean value showed significant results, except for the change in hip angle over time. For patient 3, aimed at the recreation of the step cycle in water, the angular coefficient indicated improvements within the single session for the hip and knee angles; in addition, no temporal evolution between sessions has been demonstrated. The research has some limitations, such as the use of a single camera, the variability of its position, the manual selection of the steps and the clinical subjectivity of the patients, as well as the temporal distance of the acquisitions, which could influence the results. However, in the future, objective data on kinematics in the water of patients could be used to monitor patient progress by expanding the sample of subjects involved, to support the effectiveness of hydrokinesitherapy in contexts where other motion capture systems cannot be used.File | Dimensione | Formato | |
---|---|---|---|
2024_12_Palotti_Executive_Summary.pdf
accessibile in internet per tutti
Descrizione: Executive Summary tesi
Dimensione
2.15 MB
Formato
Adobe PDF
|
2.15 MB | Adobe PDF | Visualizza/Apri |
2024_12_Palotti_Tesi.pdf
accessibile in internet per tutti
Descrizione: Testo tesi
Dimensione
38.24 MB
Formato
Adobe PDF
|
38.24 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/10589/230367