Research Projects

Intraocular Lens (IOL) implantation is a widely adopted and highly effective surgical intervention, primarily performed to restore vision following cataract extraction. Despite its clinical success, a significant number of patients experience variability in post-operative visual outcomes. These inconsistencies stem from multiple interacting factors such as anatomical differences in ocular geometry, variations in the positioning and type of implanted IOL, and the dynamic behavior of intraocular fluids. Given these complexities, conventional planning methods often fall short in predicting individualized outcomes, thereby emphasizing the urgent need for advanced computational tools to support personalized surgical planning. This project aims to bridge that gap by developing a novel, interdisciplinary computational framework that synergizes CFD with ANN. The central idea is to simulate and analyze the intraocular environment specifically the aqueous humor dynamics before and after IOL implantation, and then use this simulation data to train predictive machine learning models capable of estimating postoperative visual performance. The CFD component of the study will focus on the biophysical simulation of aqueous humor flow within the anterior chamber of the eye. These simulations will incorporate detailed anatomical reconstructions derived from imaging data and account for various surgical scenarios including different IOL designs, placements , and anatomical variations. The aim is to quantify the effects of these variables on intraocular pressure distribution, fluid shear stress, and flow-induced forces on the lens and surrounding tissues. Such insights are critical in understanding mechanisms that may contribute to complications or suboptimal outcomes, such as posterior capsule opacification or lens misalignment. Complementing the CFD analysis, an ANN-based model will be designed and trained using a comprehensive dataset that integrates both clinical parameters and the physics-informed outputs from CFD simulations. By learning complex, nonlinear relationships between input features and post-operative outcomes, the ANN will serve as a predictive tool capable of providing patient-specific forecasts. This data-driven layer not only enhances the interpretability of CFD simulations but also enables generalization across diverse patient profiles. The culmination of this research will be a hybrid computational platform that offers ophthalmic surgeons a robust decision-support tool for pre-surgical assessment and personalized IOL selection. This integration of physics-based modeling and artificial intelligence represents a paradigm shift in how surgical planning for IOL implantation is approached. The project promises not only academic contributions in terms of methodology and understanding of ocular biomechanics but also practical translational impact in real-world ophthalmic practice.