Artificial intelligence for surgical outcome prediction in glaucoma: a systematic review

Area of Science:

• Ophthalmology
• Medical Informatics
• Artificial Intelligence

Background:

• Glaucoma surgery outcomes are complex to predict.
• Rising glaucoma prevalence necessitates improved surgical outcome prediction.
• Artificial intelligence (AI) offers potential for enhanced clinical decision-making.

Purpose of the Study:

• To systematically review evidence on AI for predicting glaucoma surgery outcomes.
• To evaluate the performance of AI models compared to traditional methods.

Main Methods:

• Systematic literature search across major databases (Medline, Embase, Web of Science, Scopus).
• Inclusion of studies applying AI models to glaucoma surgery outcome prediction.
• Analysis of six studies involving 4,630 surgeries.

Main Results:

• AI models consistently outperformed traditional statistical approaches.
• The best AI model achieved 87.5% accuracy in outcome prediction.
• Key predictors included demographics, systemic health, and ophthalmic parameters.

Conclusions:

• AI demonstrates significant potential for predicting glaucoma surgery outcomes.
• Limitations include lack of external validation and standardized success metrics.
• Future research requires prospective, multicenter studies and standardized data for clinical utility.