Utilization Of Artificial Intelligence And Machine  Learning To Develop Algorithms For A Novel Laser Therapy To Treat Progressive Presbyopia

To demonstrate a Digital Twin Human Eye to predict presbyopia outcomes for different stages of presbyopia following Laser Scleral Microporation (LSM) therapy.  

Small Business Innovation Research (SBIR) grant application experimental study 

A 3D Digital Twin (DT) of the human eye was developed using a Finite Element Model (FEM) incorporating published human data and individual patient data.  This model was used to assess the dosage algorithm developed using artificial intelligence to deliver progressive doses of LSM from Stage I to Stage V of presbyopia.  Twelve eyes who underwent virtual simulations of progressive doses of LSM to recover Dynamic Range of Focus (DRoF) are presented. Predicted LogMAR Distance-Corrected Near Visual Acuity (DCNVA) was calculated from using a proprietary algorithm.  

The DT predicted results that varied by age, pore dose (9, 16, 25, 36, or 49 pores), and pore placement.  The effect of LSM was most powerful at older ages. For a 40 y.o., a 49-pore treatment resulted in a minor 9.9% increase in DCNVA.  For a 50 y.o., the 49-pore treatment improved central optical power by 16%. Increasing numbers of pores increased the improvement in DCNVA; for example, 25 pores given to a 50 y.o. provided a 12% increase in DCNVA versus a 15% increase for 49 pores at the same stage. Pores placed nearer to the limbus resulted in greater changes to DCNVA than those placed centrally in zone 2. Pore dosage matched DRoF recovery for stages III-V.  

The human eye digital twin correctly predicted, monitored, and optimized human eye processes, offering an opportunity to provide individualized precision medicine. The DT was able to predict treatment algorithms for presbyopic patients, which varied with developmental stages.