A Finite Element Approach Towards Patient-Specific Biomechanical Modeling of Keratoconus

Keratoconus (KC) is a progressive corneal disorder characterized by corneal thinning, weakening, and conical deformation, which severely impairs vision. Accurate biomechanical modeling is essential for understanding the progression of KC and developing effective, personalized treatment strategies. Neglecting the biomechanical properties of the cornea in treatments such as corneal crosslinking (CXL) and refractive surgeries like LASIK can lead to unintended optical outcomes, deviating from expected corneal curvature stabilization. This thesis advances patient-specific finite element (FE) modeling of KC by incorporating corneal softening behavior into the simulations. The primary objective is to detect the KC region on the patient geometry and simulate the softening in the detected region in both mild and severe KC patients. Finally, evaluate the accuracy of the models by comparing the simulation results with actual patient corneal tomography data. Building on an existing healthy corneal model, this work modifies the material law to represent the biomechanical weakening in KC affected regions. Three studies were conducted to address limitations in existing KC modeling approaches. The first study focused on identifying the optimal degree of softening and testing stiffness reductions of 20%, 40%, and 60% in the KC weak zone. Following that, in the second study, evaluated different softening functions: uniform, linear, and exponential and concluding that a 20% exponential reduction provided the best fit, particularly in severe KC cases. The third study examined the effects of weak zone parameters, such as radius and decay constant, and found the model to be less sensitive to these variations when optimal softening was applied. The results show that 20% exponential softening yields the most accurate simulations for severe KC, though it may overestimate degradation in mild cases, suggesting a need for more personalized softening parameters. Despite certain limitations, such as the absence of in-vivo biomechanical data, the developed FE models effectively simulate KC softening behavior. Future work should focus on refining weak zone parameters based on biomechanical test data from KC corneas, incorporating depth-dependent stiffness variations, improving material properties, and validating the model with more KC patient datasets. This research provides a foundation for more precise KC simulations, with potential applications in developing personalized treatment strategies.