Transcatheter Aortic Valve Implantation (TAVI) is a minimally invasive procedure commonly used for the treatment of severe aortic stenosis. The procedural outcome depends on the mechanical interaction between the stent frame and the surrounding aortic root. Excessive local loading may induce tissue injury or conduction disturbances, whereas insufficient contact may compromise anchoring and sealing. In this context, in silico medicine represents a valuable decision-making tool for clinicians: Finite Element Analysis (FEA) has become a well-established tool to investigate post-TAVI device–tissue interactions. However, predictive reliability is limited by structural uncertainties, particularly aortic wall thickness and tissue mechanical properties, which are often unknown in vivo. This study implements a Design of Experiments (DOE) framework combined with Uncertainty Quantification (UQ) and Sensitivity Analysis (SA) to evaluate the impact of these uncertainties in patient-specific TAVI simulations. 49 configurations were generated by varying aortic wall thickness and third-order Yeoh hyperelastic material sets. Mechanical outcomes in a Region of Interest below the membranous septum were assessed via von Mises stress variation (∆σV M ) and the Contact Pressure Index (CPI). Results demonstrated a clear hierarchy: wall thickness was the primary driver of mechanical stress, with a 60% increase in ∆σV M observed when reducing thickness from 2.83 mm to 1.98 mm, consistent with Laplace’s Law. Material properties had a secondary influence, while CPI remained largely stable, indicating that device sealing depends on global anatomy rather than local tissue stiffness. Material Set 3 provided a robust approximation of pathological tissue and ensured numerical stability. Overall, these findings emphasize that, for patient-specific TAVI simulations, accurate geometric characterization in terms of aortic wall thickness is more critical than refinement of material parameters in regions of clinical interest.
La procedura Transcatheter Aortic Valve Implantation (TAVI) è una procedura minimamente invasiva comunemente utilizzata per il trattamento della stenosi aortica. L’esito procedurale dipende dall’interazione meccanica tra il telaio dello stent e il tessuto aortico circostante: eccessivi carichi locali possono indurre lesioni del tessuto o disturbi della conduzione, mentre un contatto insufficiente può compromettere l’ancoraggio e la tenuta del dispositivo. In questo contesto, la medicina in silico rappresenta uno strumento decisionale fondamentale per i clinici: l’Analisi agli Elementi Finiti (FEA) è ormai uno strumento consolidato per indagare le interazioni tra il dispositivo e il tessuto a seguito della procedura TAVI. Tuttavia, l’affidabilità predittiva è limitata da incertezze strutturali, in particolare riguardanti lo spessore della parete aortica e le proprietà meccaniche del tessuto, spesso sconosciute in vivo. Questo studio implementa un approccio basato su Design of Experiments (DOE) combinato con Uncertainty Quantification (UQ) e Sensitivity Analysis (SA) per valutare l’impatto di tali incertezze in simulazioni TAVI paziente-specifiche. Sono state generate 49 configurazioni variando lo spessore della parete aortica e i set di materiali iperelastici di tipo Yeoh di terzo ordine. Gli esiti meccanici nella Regione di Interesse al di sotto del setto membranoso sono stati valutati tramite la variazione dello stress di von Mises (∆σV M ) e l’Indice di Pressione di Contatto (CPI). I risultati hanno evidenziato una chiara gerarchia: lo spessore della parete è il principale fattore che influenza lo stress meccanico, con un aumento del 60% di ∆σV M passando da 2.83 mm a 1.98 mm, in accordo con la legge di Laplace. Le proprietà del materiale hanno un’influenza secondaria, mentre il CPI rimane stabile, indicando che l’aderenza del dispositivo dipende più dall’anatomia globale che dalla rigidità locale del tessuto. Il Materiale Set 3 ha fornito una robusta approssimazione del tessuto patologico, garantendo stabilità numerica. Questi risultati sottolineano che, per simulazioni TAVI paziente-specifiche, una caratterizzazione geometrica accurata dello spessore aortico è più critica della definizione dei parametri dei materiali nelle regioni di interesse clinico.
Uncertainty quantification of TAVI simulations
BRAMANTE, CARLOTTA
2024/2025
Abstract
Transcatheter Aortic Valve Implantation (TAVI) is a minimally invasive procedure commonly used for the treatment of severe aortic stenosis. The procedural outcome depends on the mechanical interaction between the stent frame and the surrounding aortic root. Excessive local loading may induce tissue injury or conduction disturbances, whereas insufficient contact may compromise anchoring and sealing. In this context, in silico medicine represents a valuable decision-making tool for clinicians: Finite Element Analysis (FEA) has become a well-established tool to investigate post-TAVI device–tissue interactions. However, predictive reliability is limited by structural uncertainties, particularly aortic wall thickness and tissue mechanical properties, which are often unknown in vivo. This study implements a Design of Experiments (DOE) framework combined with Uncertainty Quantification (UQ) and Sensitivity Analysis (SA) to evaluate the impact of these uncertainties in patient-specific TAVI simulations. 49 configurations were generated by varying aortic wall thickness and third-order Yeoh hyperelastic material sets. Mechanical outcomes in a Region of Interest below the membranous septum were assessed via von Mises stress variation (∆σV M ) and the Contact Pressure Index (CPI). Results demonstrated a clear hierarchy: wall thickness was the primary driver of mechanical stress, with a 60% increase in ∆σV M observed when reducing thickness from 2.83 mm to 1.98 mm, consistent with Laplace’s Law. Material properties had a secondary influence, while CPI remained largely stable, indicating that device sealing depends on global anatomy rather than local tissue stiffness. Material Set 3 provided a robust approximation of pathological tissue and ensured numerical stability. Overall, these findings emphasize that, for patient-specific TAVI simulations, accurate geometric characterization in terms of aortic wall thickness is more critical than refinement of material parameters in regions of clinical interest.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_3_Bramante_Tesi.pdf
non accessibile
Descrizione: Testo della tesi
Dimensione
19.29 MB
Formato
Adobe PDF
|
19.29 MB | Adobe PDF | Visualizza/Apri |
|
2026_3_Bramante_ExecutiveSummary.pdf
non accessibile
Descrizione: Testo Executive Summary
Dimensione
3.62 MB
Formato
Adobe PDF
|
3.62 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/10589/252669