Application Of Machine Learning Algorithms To Improve The Accuracy Of Laser Vision Correction Results

To explore the potential of machine learning (ML) algorithms for improving the precision of laser refractive surgery outcomes.

Department of Ophthalmology, Faculty of Continuing Professional Education and Professional Retraining, Russian National Research Medical University named after N.I. Pirogov, Moscow, Russia

This study investigates the applicability of machine learning algorithms to enhance the accuracy of laser refractive surgery. For processing multimodal data, distinct ML methods are employed: convolutional neural networks (CNNs) for image analysis, regression analysis for temporal data processing and value prediction, deep learning for text-based data.

Machine learning can elevate refractive surgery outcomes through the following mechanisms: improved detection of subclinical risk factors for ectasia through deep feature extraction from multimodal images; optimization of ablation profiles beyond the classical Mannerlin formula using physics-informed neural networks; personalization of postoperative regimens through continuous outcome prediction models.

The accuracy of outcomes in laser refractive surgery depends on a multitude of factors. Machine learning aids in analyzing large sets of variables to address this challenge.

The findings of this work could be implemented as a medical information system capable of aggregating and integrating multimodal diagnostic data, suggest the volume of refractive surgery and predict the outcome (decision support system).