The curvature-adaptive voxel Monte Carlo (CAVMC) method-based photothermal model for customized retinal laser surgery

Laser photothermal therapy is widely used in ophthalmology to treat retinal diseases such as diabetic retinopathy and age-related macular degeneration. Precise control of laser energy absorbed by the retinal pigment epithelium (RPE) and blood vessels is crucial to avoid insufficient or excessive treatment. To overcome the limitations of traditional Monte Carlo (MC) methods including Voxel-based (VMC) and Tetrahedral-based (TMC) MC, a new Curvature-Adaptive Voxel-based Monte Carlo (CAVMC) method is proposed. The CAVMC introduces artificial surfaces aligned with curved interfaces to improve the accuracy of light transport simulations, thereby achieving near-perfect agreement with TMC results while reducing computational costs by 70 %. Simulations of various retinal laser treatment protocols, including Panretinal Photocoagulation (PRP), Transpupillary Thermotherapy (TTT), Micropulse Laser Therapy (MLT), and Selective Retina Therapy (SRT), demonstrated that CAVMC reduces thermal prediction errors by up to 25 % compared to VMC. This method offers a powerful tool for optimizing retinal laser surgery, balancing high accuracy with computational efficiency, and holds promise for broader biomedical applications.