Computational models have been increasingly utilized to evaluate patient conditions, to determine the severity of a pathology or evaluate and test different solutions to apply. These models ensure the repeatability of the evaluations that are carried out and often offer insights that would not be easily obtained from patients, such as different levels of stresses and deformations obtained as a consequence of the chosen solutions applied. The routine application of these models in clinical practice is however limited by the presence of approximations, which are necessary either to contain computational costs or due to lack of digital integration of data obtainable from different clinical analyses (i.e. motion capture analyses and radiology exams). A general workflow has therefore been developed to address clinical questions, to offer correct and reliable insights, and consequently to properly define risks and tailor solutions to patient issues. Each detail of the workflow can be then tailored to the specific case, accounting for the typology of data available, clinical requirements, and patient-specific variabilities. This allows for the integration of non-correlated data obtainable from different sources into a unified numerical model. Image-based models will be utilized, ensuring that the imaging datasets possess the characteristics required to calculate patient-specific biomechanical properties and adopting proper strategies to compensate possible criticalities that may emerge. The boundary conditions set may aim to answer specific clinical questions or to replicate realistic muscle loads occurring during clinically relevant movements. Once these high accuracy models have been generated, the identified patient-specific issues can be addressed with patient-specific solutions. This may result in the definition of specific parameters, such as risk factors, or in the evaluation of patient-specific designed devices and different fixation methods. The design freedom and flexibility offered by additive manufacturing may become necessary to achieve complex geometries. The PhD project therefore aims at addressing patient-specific issues and solutions in the field of orthopaedic surgery, following the entire path from the acquisition of patient images until the generation of information relevant to specific clinical issues or evaluate and design devices. Four main anatomical regions have been evaluated: (1) the lumbar spine, including the first and second lumbar vertebrae, (2) the cervical spine, specifically the second cervical vertebra, (3) the clavicle, and (4) the gleno-humeral joint, including the scapula and the humerus. These apparently different anatomical regions have specific elements in common, as their numerical modelling can benefit from the structural imaging of hard tissues and the presence of realistic dynamic and kinematic parameters which function as boundary conditions. All these information can be obtained from different sources and implemented through different tools, and their integration requires different choices strictly connected to the clinically relevant outcomes. Furthermore, their peculiarities allow for the highlighting of different sections of the general workflow.
Modelli computazionali sono sempre più utilizzati per valutare le condizioni di pazienti, in modo da determinare la gravità di una patologia o valutare e testare l'effetto di diverse soluzioni da applicare. Questi modelli garantiscono la ripetibilità delle analisi svolte e spesso offrono informazioni aggiuntive che non risultano facilmente acquisibili dai pazienti stessi, come per esempio diversi livelli di sforzi e deformazioni ottenuti in base alle soluzioni introdotte. L'applicazione di questi modelli alla pratica clinica di routine è però limitata dalla presenza di approssimazioni, le quali risultano necessarie per limitare i costi computazionali o a causa di una mancante integrazione digitale delle informazioni acquisibili tramite analisi cliniche (es. analisi del movimento ed esami radiologici). Un processo generalizzato è stato quindi sviluppato per poter affrontare problematiche cliniche, per offrire informazioni corrette ed affidabili, e di conseguenza per definire rischi e adattare soluzioni alle necessità del paziente. Ogni dettaglio del processo può essere adattato ad ogni specifico caso, in base alla tipologia di dato disponibile, alle richieste cliniche, e alle specificità del paziente. Questo permette dunque di integrare dati non correlati provenienti da diverse fonti in un unico modello numerico. Verranno utilizzati modelli basati su immagini, i quali dovranno possedere le necessarie caratteristiche per poter stimare proprietà biomeccaniche paziente-specifiche; strategie appropriate verranno adottate per compensare possibili problematicità in caso contrario. Le condizioni al contorno impostate potranno rispondere a specifiche richieste cliniche o riprodurre carichi muscolari realistici che avvengono durante movimenti ritenuti rilevanti dal punto di vista clinico. Generati questi modelli caratterizzati da elevata accuratezza, le problematiche paziente specifiche verranno approcciate nell'ottica dello sviluppo di soluzioni paziente specifiche. Questo porterà alla definizione di specifici parametri, ad esempio fattori di rischio, o all'analisi del design di impianti paziente-specifici e sistemi di fissazione. La libertà di design e flessibilità garantita dalla manifattura additiva risulterà necessaria per ottenere geometrie complesse. Questo progetto di dottorato si pone dunque come obiettivo l'analisi di problematiche e la generazione di soluzioni paziente-specifiche nell'ambito della chirurgia ortopedica, seguendo l'intero percorso dall'acquisizione delle immagini del paziente fino alla definizione di informazioni rilevanti per specifiche problematiche cliniche o per analizzare e generare impianti. Quattro regioni anatomiche sono state valutate: (1) la zona lombare della colonna vertebrale, in particolare la prima e seconda vertebra lombare, (2) la seconda vertebra cervicale, (3) la clavicola, e (4) l'articolazione gleno-omerale, la quale comprende sia la scapola che l'omero. Queste regioni anatomiche apparentemente differenti hanno specifici elementi in comune, in quanto modellazione numerica di questi segmenti può beneficiare della rappresentazione strutturale dei tessuti ossei, e della presenza di parametri dinamici e cinematici realistici, i quali vengono utilizzati come condizioni al contorno. Tutte queste informazioni possono essere acquisite da fonti diverse ed implementate mediante diversi strumenti; la loro integrazione richiede scelte modellistiche differenti, le quali sono strettamente correlati agli obiettivi considerati rilevanti clinicamente. Inoltre, le particolarità di ogni segmento permette di evidenziare diverse porzioni del processo generalizzato.
Patient-specific modelling in orthopaedics: a path to clinically consistent integration
De CET, ANNA
2025/2026
Abstract
Computational models have been increasingly utilized to evaluate patient conditions, to determine the severity of a pathology or evaluate and test different solutions to apply. These models ensure the repeatability of the evaluations that are carried out and often offer insights that would not be easily obtained from patients, such as different levels of stresses and deformations obtained as a consequence of the chosen solutions applied. The routine application of these models in clinical practice is however limited by the presence of approximations, which are necessary either to contain computational costs or due to lack of digital integration of data obtainable from different clinical analyses (i.e. motion capture analyses and radiology exams). A general workflow has therefore been developed to address clinical questions, to offer correct and reliable insights, and consequently to properly define risks and tailor solutions to patient issues. Each detail of the workflow can be then tailored to the specific case, accounting for the typology of data available, clinical requirements, and patient-specific variabilities. This allows for the integration of non-correlated data obtainable from different sources into a unified numerical model. Image-based models will be utilized, ensuring that the imaging datasets possess the characteristics required to calculate patient-specific biomechanical properties and adopting proper strategies to compensate possible criticalities that may emerge. The boundary conditions set may aim to answer specific clinical questions or to replicate realistic muscle loads occurring during clinically relevant movements. Once these high accuracy models have been generated, the identified patient-specific issues can be addressed with patient-specific solutions. This may result in the definition of specific parameters, such as risk factors, or in the evaluation of patient-specific designed devices and different fixation methods. The design freedom and flexibility offered by additive manufacturing may become necessary to achieve complex geometries. The PhD project therefore aims at addressing patient-specific issues and solutions in the field of orthopaedic surgery, following the entire path from the acquisition of patient images until the generation of information relevant to specific clinical issues or evaluate and design devices. Four main anatomical regions have been evaluated: (1) the lumbar spine, including the first and second lumbar vertebrae, (2) the cervical spine, specifically the second cervical vertebra, (3) the clavicle, and (4) the gleno-humeral joint, including the scapula and the humerus. These apparently different anatomical regions have specific elements in common, as their numerical modelling can benefit from the structural imaging of hard tissues and the presence of realistic dynamic and kinematic parameters which function as boundary conditions. All these information can be obtained from different sources and implemented through different tools, and their integration requires different choices strictly connected to the clinically relevant outcomes. Furthermore, their peculiarities allow for the highlighting of different sections of the general workflow.| File | Dimensione | Formato | |
|---|---|---|---|
|
PhD_Thesis_ACDC.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: phd thesis document
Dimensione
233.48 MB
Formato
Adobe PDF
|
233.48 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/10589/256377