Enhancing Predicted Visual Acuity After SmartSight Lenticule Extraction: Identifying Key Factors With Machine...

Area of Science:

• Ophthalmology
• Laser Vision Correction
• Biomedical Engineering

Background:

• Refractive surgery aims to improve visual acuity.
• Lenticule extraction with the SmartSight system is a common procedure.
• Predicting visual outcomes is crucial for patient management.

Purpose of the Study:

• Develop a predictive model for corrected distance visual acuity (CDVA) improvement after SmartSight lenticule extraction.
• Evaluate the influence of clinical and procedural parameters on visual acuity enhancement.
• Assess the effectiveness of various parameters in predicting visual outcomes.

Main Methods:

• Analysis of data from 1,262 eyes treated with the SmartSight system.
• Utilized regression and classification techniques with machine learning for CDVA gain prediction.
• Employed analysis of variance and principal components analysis to identify significant features.

Main Results:

• Identified three principal components explaining over 70% of data variance.
• Key factors for CDVA improvement include spot-to-track distance ratio, track distance, pulse energy, laser power, spot distance, cap thickness, and dosage.
• System settings, patient gender, and laterality were also associated with outcomes.

Conclusions:

• Low-energy asymmetric spacing in lenticule extraction with the SmartSight system is safe and effective.
• This approach enhances visual outcomes for myopic astigmatism.
• The developed model reliably predicts CDVA improvements, aiding clinical decision-making.