AI-based applications in ophthalmology: a discussion on future challenges in its use for diagnosis and treatment in the area

• DOI: https://doi.org/10.22533/at.ed.1594582426069

• Palavras-chave: ophthalmology, Artificial Intelligence; Treatment; Diagnosis.

• Keywords: ophthalmology, Artificial Intelligence; Treatment; Diagnosis.

• Abstract:

  Objective: Explore the scope of applications of artificial intelligence in ophthalmology, discussing the technical-ethical-regulatory challenges they may face. Methods: Bibliographic review through the PubMed database, 730 articles were found subsequently submitted to the criteria of the PVO strategy (research, variables and outcome), through the search strategy (Artificial intelligence) AND (Ophthalmology) AND (Diagnosis) AND (Treatment). After applying the selection criteria, 24 articles were selected to compose the present study. Review: AI programs already implemented and in experimentation are discussed, such as the IDx-dr system, an AI technology approved by the FDA to allow the diagnosis of diabetic retinopathy, the “Redcheck” platform, which in Brazil facilitates the screening of ocular changes and assists the ophthalmologist in preparing quick reports, among other highlighted artificial intelligence alternatives. Final considerations: Contact with the programs already implemented allowed us to see the potential of AI to assist ophthalmological medicine, despite having contact with the limitations already found with regard to validity and effectiveness on a large scale, certification of efficiency and ethical commitment to the implementation of AI. 
  Finally, it was concluded that there is a need to carry out more studies to provide comprehensive evidence for the safe implementation of AI, already respecting its established potential in the medical world and in ophthalmology.