Keratoconus: exploring fundamentals and future perspectives - a comprehensive systematic review.

Area of Science:

• Ophthalmology
• Medical Artificial Intelligence
• Corneal Diseases

Background:

• Artificial intelligence (AI) advancements have significantly improved keratoconus detection and management.
• AI integration in technologies like OCT and femtosecond lasers enhances treatment safety, precision, and predictability.
• AI facilitates less invasive approaches to keratoconus treatment, from contact lenses to keratoplasty.

Purpose of the Study:

• To comprehensively review keratoconus treatment modalities.
• To explore the role and application of machine learning strategies in keratoconus management.

Main Methods:

• A systematic narrative review was conducted.
• A comprehensive search of five major electronic databases (PubMed, Scopus, Web of Science, Embase, Cochrane) was performed without restrictions.
• Eligible articles were selected based on title/abstract screening and full-text review.

Main Results:

• AI shows significant promise in keratoconus diagnosis and management, including early detection and screening.
• AI aids in predicting postoperative ectasia and guiding surgical decisions, such as intracorneal ring segment implantation.
• While most studies used single algorithms, some assessed multiple algorithms for staging and management strategies.

Conclusions:

• Widespread clinical translation of AI models is essential for advancing keratoconus management.
• Future approaches leveraging AI aim to optimize visual performance for keratoconus patients.