The cornea is a transparent dome shaped tissue constituting the outer part of the eye, and which functions aim to protect the eye, and to refract and focus the light onto the retina. Its biological structure is characterized by the interaction between extracellular matrix and highly organized collagen lamellae, thus resulting in a visco-hyperelastic anisotropic tissue with a non-linear behavior. Such architecture may be liable to thinning, stiffening, bulging, and other changes due to both ocular diseases such as keratoconus, and surgeries such as refractive surgery and corneal cross-linking, hence impairing eyesight. Non-contact tonometry (NCT) is a diagnostic instrument aimed at estimating intraocular pressure (IOP) in vivo by exploiting an air pulse able to deform the cornea, which deforms downward reaching three main configurations: first applanation, highest concavity, and second applanation. Corneal deformation is a result of four interacting factors: external load (the air pulse pressure), IOP, corneal geometry, in particular its thickness, and corneal tissue intrinsic mechanical properties. It is therefore crucial to understand how each factor is related to corneal deformation, to provide both accurate IOP measurement and accurate mechanical properties estimates to discriminate healthy from pathological corneas. To date, the relationships between corneal deformation biomarkers during the NCT and each affecting factor are feasible only via numerical simulations. Currently, the relationships between Corvis ST biomarkers have been studied by various authors through clinical statistical analysis. However, there is a lack of agreement among the results, making it difficult to draw precise conclusions. This thesis aims to deepen the knowledge of intrinsic corneal mechanical properties and their relationships with the NCT deformation results, to find a methodology to estimate them in vivo, which is currently missing. The presented work has been divided into three main sections. In the first section, a clinical analysis has been performed to analyze relationships between corneal deformation biomarkers and IOP, corneal thickness, and age, hence contributing to the knowledge of in vivo corneal behavior. Then, in the second section, a numerical analysis has been carried out to study the relationships between corneal deformation biomarkers and IOP, and intrinsic corneal mechanical properties. Results obtained in this second section have been exploited in the third section to train an artificial neural network aimed to predict IOP and corneal intrinsic mechanical properties based on the corneal deformation biomarkers. Results obtained in the first two sections confirm that corneal deformation during the NCT is affected by the interaction of different factors. Results of the third section highlight the capability of our ANN model to predict IOP and corneal intrinsic mechanical properties based on corneal deformation biomarkers. This thesis has provided a deep understanding of intrinsic corneal mechanical properties and their relationships with the NCT deformation results, together with a possible solution to estimate them in vivo.
La cornea è un tessuto trasparente di forma sferica che costituisce la parte più esterna dell’occhio, le cui funzioni principali sono la protezione dell’occhio, la rifrazione e la messa a fuoco della luce sulla retina. La sua struttura biologica è caratterizzata dall’interazione tra la matrice extracellulare e le lamelle di collagene altamente organizzate, così da generare un tessuto visco-iperelastico anisotropo con comportamento non lineare. La struttura della cornea può essere soggetta ad assottigliamento, irrigidimento, rigonfiamento e altre modifiche causate sia da patologie oculari come il cheratocono, che da interventi chirurgici come la chirurgia refrattiva e il cross-linking corneale, portando così a una perdita della vista. La tonometria senza contatto (NCT) è uno strumento diagnostico in grado di predire la pressione intraoculare (PIO) in vivo sfruttando l’interazione tra un getto d’aria e la cornea, la quale si deforma raggiungendo tre configurazioni principali: prima applanazione, concavità massima e seconda applanazione. La deformazione della cornea è data dall’interazione di quattro fattori: il carico esterno (pressione del getto d’aria), la PIO, la geometria della cornea, in particolare il suo spessore, e le proprietà meccaniche intrinseche della cornea. E’ quindi essenziale capire come ognuno di questi fattori influenzi la deformazione della cornea, in modo da fornire sia un valore di PIO accurato sia un’accurata stima delle proprietà meccaniche, così da distinguere cornee sane da cornee patologiche. Ad oggi, relazionare la deformazione della cornea durante la NCT con ognuno dei fattori è possibile solo tramite simulazioni numeriche. Attualmente, le relazioni tra i biomarcatori del Corvis ST sono state studiate da vari autori tramite analisi statistiche cliniche. Tuttavia, c’è un’assenza di concordanza tra i risultati, rendendo dunque difficile ottenere conclusioni precise. Questa tesi mira ad approfondire la conoscenza delle proprietà meccaniche intrinseche della cornea e le loro relazioni con i risultati di deformazione ottenuti mediante la NCT, in modo da trovare una metodologia per stimarle in vivo, la quale è ancora assente. Il seguente lavoro è stato suddiviso in tre sezioni principali. Nella prima sezione, è stata eseguita un’analisi clinica per studiare le relazioni tra i biomarcatori di deformazione della cornea e la PIO, lo spessore della cornea e l’età, così da approfondire il comportamento della cornea in vivo. Successivamente, nella seconda sezione è stata eseguita un’analisi numerica per studiare le relazioni tra i biomarcatori di deformazione della cornea e la PIO e le proprietà meccaniche intrinseche della cornea. I risultati ottenuti in questa seconda sezione sono stati poi sfruttati nella terza sezione per allenare una rete neurale artificiale in grado di predire la PIO e le proprietà meccaniche intrinseche della cornea a partire dai biomarcatori di deformazione della cornea. I risultati ottenuti nelle prime due sezioni confermano che la deformazione della cornea durante la NCT è influenzata dall’interazione di vari fattori. I risultati della terza sezione sottolineano la capacità di predire la PIO e le proprietà meccaniche intrinseche della cornea a partire dai biomarcatori di deformazione della cornea. Questa tesi ha contribuito a fornire una conoscenza approfondita delle proprietà meccaniche intrinseche della cornea e delle loro relazioni con i risultati di deformazione della NCT, insieme ad una possibile soluzione per stimarle in vivo.
Statistical analysis of Corvis ST biomarkers in a Montecarlo simulation of the Non-Contact Tonometry
ALLIEVI, ELEONORA
2023/2024
Abstract
The cornea is a transparent dome shaped tissue constituting the outer part of the eye, and which functions aim to protect the eye, and to refract and focus the light onto the retina. Its biological structure is characterized by the interaction between extracellular matrix and highly organized collagen lamellae, thus resulting in a visco-hyperelastic anisotropic tissue with a non-linear behavior. Such architecture may be liable to thinning, stiffening, bulging, and other changes due to both ocular diseases such as keratoconus, and surgeries such as refractive surgery and corneal cross-linking, hence impairing eyesight. Non-contact tonometry (NCT) is a diagnostic instrument aimed at estimating intraocular pressure (IOP) in vivo by exploiting an air pulse able to deform the cornea, which deforms downward reaching three main configurations: first applanation, highest concavity, and second applanation. Corneal deformation is a result of four interacting factors: external load (the air pulse pressure), IOP, corneal geometry, in particular its thickness, and corneal tissue intrinsic mechanical properties. It is therefore crucial to understand how each factor is related to corneal deformation, to provide both accurate IOP measurement and accurate mechanical properties estimates to discriminate healthy from pathological corneas. To date, the relationships between corneal deformation biomarkers during the NCT and each affecting factor are feasible only via numerical simulations. Currently, the relationships between Corvis ST biomarkers have been studied by various authors through clinical statistical analysis. However, there is a lack of agreement among the results, making it difficult to draw precise conclusions. This thesis aims to deepen the knowledge of intrinsic corneal mechanical properties and their relationships with the NCT deformation results, to find a methodology to estimate them in vivo, which is currently missing. The presented work has been divided into three main sections. In the first section, a clinical analysis has been performed to analyze relationships between corneal deformation biomarkers and IOP, corneal thickness, and age, hence contributing to the knowledge of in vivo corneal behavior. Then, in the second section, a numerical analysis has been carried out to study the relationships between corneal deformation biomarkers and IOP, and intrinsic corneal mechanical properties. Results obtained in this second section have been exploited in the third section to train an artificial neural network aimed to predict IOP and corneal intrinsic mechanical properties based on the corneal deformation biomarkers. Results obtained in the first two sections confirm that corneal deformation during the NCT is affected by the interaction of different factors. Results of the third section highlight the capability of our ANN model to predict IOP and corneal intrinsic mechanical properties based on corneal deformation biomarkers. This thesis has provided a deep understanding of intrinsic corneal mechanical properties and their relationships with the NCT deformation results, together with a possible solution to estimate them in vivo.File | Dimensione | Formato | |
---|---|---|---|
2024_12_Allievi_Tesi.pdf
non accessibile
Descrizione: Testo della tesi
Dimensione
9.47 MB
Formato
Adobe PDF
|
9.47 MB | Adobe PDF | Visualizza/Apri |
2024_12_Allievi_Executive_Summary.pdf
non accessibile
Descrizione: Testo dell'executive summary
Dimensione
1.06 MB
Formato
Adobe PDF
|
1.06 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/10589/230929