Preoperative assessment for refractive surgery: machine learning

Machine learning is revolutionizing preoperative evaluation in refractive surgery.

Introduction

Machine learning is a general technique for finding suitable parameters or functions to classify input data from large amounts of training data.

Methodologies

Several methodologies have been recommended:

• Support Vector Machines (SVM) — classification by hyperplanes
• Decision trees — structured predictive models
• Neural networks — models inspired by the human brain

Deep learning

Deep learning deals with the training of multi-layer artificial neural networks. Convolutional neural networks have achieved impressive results in image classification in ophthalmology and many scientific fields.

Application in ophthalmology

In refractive surgery, these technologies allow:

• Detection of contraindications (frustic keratoconus)
• Prediction of visual results postoperative
• Personalization of choice technique (LASIK, PRK, SMILE)
• Improved security of the intervention