Despite the growing use of instrumental motion analysis in clinical research and practice, existing experimental protocols are often too generic to capture condition-specific measurements, including possible compensation patterns, or too specific to one pathology, which limits their broader applicability. This thesis aims to address this issue by designing and validating a standardised yet flexible 3D motion analysis framework that can be adapted to different spinal disorders. The proposed protocol enables the functional assessment of trunk and lower limb kinematics in both paediatric and adult populations. The protocol was applied to two distinct clinical cohorts: adolescent with idiopathic scoliosis (AIS), who were treated with either vertebral body tethering (VBT) or posterior spinal fusion (PSF), and adult patients who underwent vertebrectomy for spinal tumours involving the cervical, thoracic or lumbar regions. This cross-pathology approach is demonstratimg the potential of a unified kinematic protocol to capture compensatory strategies and segmental contributions, regardless of the underlying spinal condition. Multiple motor tasks were evaluated, namely level walking, and elementary excercises such as flexion–extension, lateral bending and axial rotation, to characterise compensatory strategies and segmental contributions. A custom Python-based pipeline was developed to automate the extraction, interpolation and visualisation of clinically relevant kinematic variables. This ensured objective and reproducible analysis, and laid the groundwork for future studies on larger datasets. The repeatability of the extracted kinematic variables was evaluated using repeated sessions from a healthy control subject. All joint-level variables showed excellent repeatability, with intraclass correlation coefficients (ICC) exceeding 0.75 and root mean square errors (RMSE) below 5°, thereby confirming the robustness of the protocol across tasks and movement planes. Ground reaction forces were acquired and relevant joint moments were analysed using an established processing pipeline. However, as there were no significant differences across experimental conditions in either group, the current work focuses exclusively on the kinematic results. Spatiotemporal parameters, compared to controls, revealed a generalized gait impairment in all pathological groups, with reduced walking speed compared to controls and decreased cadence, despite relatively stable stride times. In the AIS cohort, VBT patients showed preserved axial rotation during gait , but also persistent asymmetries in thoracic–pelvic coordination and frontal-plane deviation during flexion–extension, likely due to incomplete correction of the scoliotic curvature. PSF patients showed more symmetrical but more constrained movement patterns, especially in the thoracic region. In the spinal tumor patients, the surgical impact on trunk movement was shown to vary depending on tumor location and fusion level. Cervical patients relied more heavily on hip motion as expected during sagittal tasks , while thoracic patients exhibited persistent upper trunk rigidity, particularly in axial rotation. Lumbar patients demonstrated an decreased of lumbo-pelvic index, indicating greater pelvic compensation and diminished lumbar segmental contribution during flexion–extension. Overall, this work introduces a rigorous methodological and computational framework for cross-pathology spinal motion analysis. By standardizing the protocol and automating data processing, it enables clinically meaningful, quantitative assessments of post-surgical function that extend beyond traditional imaging and patient-reported outcomes. Future studies should integrate kinetic and electromyographic data to further elucidate neuromuscular adaptations and support personalized rehabilitation planning.
Nonostante il crescente utilizzo dell’analisi strumentale del movimento nella ricerca clinica e nella pratica medica, i protocolli sperimentali attualmente disponibili risultano spesso troppo generici per cogliere misurazioni specifiche della condizione, incluse eventuali strategie compensatorie, oppure troppo specifici per una singola patologia, limitandone così l’applicabilità trasversale. Questa tesi si propone di superare tali limitazioni mediante la progettazione e la validazione di un protocollo standardizzato ma flessibile per l’analisi tridimensionale del movimento, adattabile a diverse patologie della colonna vertebrale. Il protocollo consente la valutazione funzionale della cinematica del tronco e degli arti inferiori sia in popolazioni pediatriche che adulte. È stato applicato a due coorti cliniche distinte: adolescenti affetti da scoliosi idiopatica (AIS), trattati mediante tethering vertebrale anteriore (VBT) o artrodesi posteriore (PSF), e pazienti adulti sottoposti a vertebrectomia per tumori spinali localizzati a livello cervicale, toracico o lombare. Questo approccio trasversale ha dimostrato il potenziale di un protocollo cinematico unificato nella rilevazione delle strategie compensatorie e delle contribuzioni segmentali, indipendentemente dalla patologia di base. Sono stati valutati diversi compiti motori, tra cui cammino su superficie piana e movimenti elementari come flesso-estensione, bending laterale e rotazione assiale, al fine di caratterizzare le strategie compensatorie e le contribuzioni dei singoli segmenti. È stata sviluppata una pipeline personalizzata in Python per automatizzare l’estrazione, l’interpolazione e la visualizzazione delle variabili cinematiche clinicamente rilevanti. Ciò ha garantito un’analisi oggettiva e riproducibile, gettando le basi per futuri studi su campioni più ampi. La ripetibilità delle variabili cinematiche estratte è stata valutata tramite sessioni ripetute su un soggetto sano. Tutte le variabili articolari hanno mostrato un'eccellente ripetibilità, con coefficienti di correlazione intraclasse (ICC) superiori a 0.75 e errori quadratici medi (RMSE) inferiori a 5°, confermando la robustezza del protocollo attraverso diversi compiti e piani di movimento. Le forze di reazione al suolo e i momenti articolari sono stati acquisiti e analizzati tramite una pipeline consolidata; tuttavia, in assenza di differenze significative tra le condizioni sperimentali in entrambi i gruppi, il presente lavoro si concentra esclusivamente sui risultati cinematici. I parametri spaziotemporali, confrontati con i controlli, hanno rivelato un’alterazione generalizzata del cammino in tutti i gruppi patologici, con una riduzione della velocità e della cadenza, a fronte di tempi del passo relativamente stabili. Nella coorte AIS, i pazienti VBT hanno mostrato una rotazione assiale preservata durante il cammino, ma anche persistenti asimmetrie nella coordinazione toraco-pelvica e deviazioni sul piano frontale durante la flesso-estensione, probabilmente dovute a una correzione incompleta della curva scoliotica. I pazienti PSF hanno evidenziato movimenti più simmetrici ma anche più limitati, soprattutto nella regione toracica. Nei pazienti oncologici sottoposti a vertebrectomia, l’impatto chirurgico sul movimento del tronco è risultato variabile in funzione della sede del tumore e del livello di fusione. I pazienti con resezione cervicale hanno mostrato una maggiore dipendenza dal movimento dell’anca nei compiti sagittali, come atteso; quelli con resezione toracica hanno mantenuto una rigidità persistente del tronco superiore, in particolare in rotazione assiale; mentre i pazienti lombari hanno evidenziato una riduzione dell’indice lumbo-pelvico, suggerendo una maggiore compensazione pelvica e una minore contribuzione segmentale lombare durante la flesso-estensione. Nel complesso, questo lavoro introduce un solido framework metodologico e computazionale per l’analisi del movimento spinale in presenza di diverse patologie. La standardizzazione del protocollo e l’automazione del processo analitico permettono valutazioni quantitative, clinicamente rilevanti, della funzionalità post-operatoria, superando i limiti delle sole immagini radiologiche o dei questionari soggettivi. Studi futuri dovranno integrare dati cinematici, dinamici ed elettromiografici per comprendere meglio le strategie neuromuscolari adattive e supportare la pianificazione di percorsi riabilitativi personalizzati.
Instrumental gait-analysis of the trunk before and after spinal surgery: development of an experimental protocol for primary bone tumors and adolescencent idiopathic scoliosis
Bressan, Francesca
2024/2025
Abstract
Despite the growing use of instrumental motion analysis in clinical research and practice, existing experimental protocols are often too generic to capture condition-specific measurements, including possible compensation patterns, or too specific to one pathology, which limits their broader applicability. This thesis aims to address this issue by designing and validating a standardised yet flexible 3D motion analysis framework that can be adapted to different spinal disorders. The proposed protocol enables the functional assessment of trunk and lower limb kinematics in both paediatric and adult populations. The protocol was applied to two distinct clinical cohorts: adolescent with idiopathic scoliosis (AIS), who were treated with either vertebral body tethering (VBT) or posterior spinal fusion (PSF), and adult patients who underwent vertebrectomy for spinal tumours involving the cervical, thoracic or lumbar regions. This cross-pathology approach is demonstratimg the potential of a unified kinematic protocol to capture compensatory strategies and segmental contributions, regardless of the underlying spinal condition. Multiple motor tasks were evaluated, namely level walking, and elementary excercises such as flexion–extension, lateral bending and axial rotation, to characterise compensatory strategies and segmental contributions. A custom Python-based pipeline was developed to automate the extraction, interpolation and visualisation of clinically relevant kinematic variables. This ensured objective and reproducible analysis, and laid the groundwork for future studies on larger datasets. The repeatability of the extracted kinematic variables was evaluated using repeated sessions from a healthy control subject. All joint-level variables showed excellent repeatability, with intraclass correlation coefficients (ICC) exceeding 0.75 and root mean square errors (RMSE) below 5°, thereby confirming the robustness of the protocol across tasks and movement planes. Ground reaction forces were acquired and relevant joint moments were analysed using an established processing pipeline. However, as there were no significant differences across experimental conditions in either group, the current work focuses exclusively on the kinematic results. Spatiotemporal parameters, compared to controls, revealed a generalized gait impairment in all pathological groups, with reduced walking speed compared to controls and decreased cadence, despite relatively stable stride times. In the AIS cohort, VBT patients showed preserved axial rotation during gait , but also persistent asymmetries in thoracic–pelvic coordination and frontal-plane deviation during flexion–extension, likely due to incomplete correction of the scoliotic curvature. PSF patients showed more symmetrical but more constrained movement patterns, especially in the thoracic region. In the spinal tumor patients, the surgical impact on trunk movement was shown to vary depending on tumor location and fusion level. Cervical patients relied more heavily on hip motion as expected during sagittal tasks , while thoracic patients exhibited persistent upper trunk rigidity, particularly in axial rotation. Lumbar patients demonstrated an decreased of lumbo-pelvic index, indicating greater pelvic compensation and diminished lumbar segmental contribution during flexion–extension. Overall, this work introduces a rigorous methodological and computational framework for cross-pathology spinal motion analysis. By standardizing the protocol and automating data processing, it enables clinically meaningful, quantitative assessments of post-surgical function that extend beyond traditional imaging and patient-reported outcomes. Future studies should integrate kinetic and electromyographic data to further elucidate neuromuscular adaptations and support personalized rehabilitation planning.File | Dimensione | Formato | |
---|---|---|---|
2025_07_Bressan.pdf
accessibile in internet per tutti
Dimensione
13.88 MB
Formato
Adobe PDF
|
13.88 MB | Adobe PDF | Visualizza/Apri |
2025_07_Bressan_executive summary.pdf
accessibile in internet per tutti
Dimensione
599.61 kB
Formato
Adobe PDF
|
599.61 kB | 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/10589/240973