Applying Machine Learning Techniques in Nomogram Prediction and Analysis for SMILE Treatment.

Area of Science:

• Ophthalmology
• Artificial Intelligence
• Refractive Surgery

Background:

• Small incision lenticule extraction (SMILE) is a popular refractive surgery procedure.
• Accurate nomograms are crucial for optimizing visual outcomes in SMILE.
• Surgeon-set nomograms may have limitations in precision.

Purpose of the Study:

• To compare the efficacy and predictability of a machine learning-based nomogram versus a surgeon-set nomogram for SMILE surgery.
• To analyze visual outcomes, including safety, efficacy, predictability, and spherical equivalent (SE) correction.

Main Methods:

• A prospective, comparative clinical study involving 865 ideal SMILE cases.
• Machine learning model trained on experienced surgeon's data to predict nomograms.
• Comparison of visual outcomes between surgeon-set and machine learning-predicted nomograms.

Main Results:

• Machine learning group showed a significantly higher efficacy index (1.48 ± 1.08) than the surgeon group (1.3 ± 0.27).
• 93% of eyes in the machine learning group achieved ±0.50 D accuracy, versus 83% in the surgeon group.
• Both groups demonstrated high predictability within ±1.00 D (96-98%), with minor differences in SE correction error.

Conclusions:

• Machine learning demonstrates comparable safety and predictability to surgeons for SMILE nomograms.
• AI-powered nomograms significantly enhance surgical efficacy and visual outcomes.
• Machine learning shows potential for improving refractive surgery precision, especially for complex cases.