Patient Perspectives on Artificial Intelligence-Based Diabetic Retinopathy Screening at an Urban US Medical Center

Zainab Rustam,1,* Yangyiran Xie,2,* Jose Amezcua Moreno,1 Diep Tran,1 Gina Zhu,1 Sophia E Yu,1 Aregnazan Sandrosyan,3 Cindy X Cai1,4 1Wilmer Eye Institute, Johns Hopkins School of Medicine, Baltimore, MD, USA; 2Vanderbilt Eye Institute, Vanderbilt University Medical Center, Nashville, TN, USA; 3Paul Foster School of Medicine, Texas Tech University Health Sciences Center El Paso, El Paso, TX, USA; 4Biomedical Informatics and Data Science, Division of General Internal Medicine, Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, MD, USA*These authors contributed equally to this workCorrespondence: Cindy X Cai, Wilmer Eye Institute, Johns Hopkins School of Medicine, 1800 Orleans Street, Room 711, Baltimore, MD, 21287, USA, Tel +14105022789, Email [email protected]: Fully autonomous artificial intelligence (AI) can improve access to diabetic retinopathy (DR) screening. In this study, we explore patient perspectives on the use of AI in DR screening.Methods: Adults with diabetes at an urban academic medical center were recruited to undergo imaging with a handheld AI fundus camera and respond to a survey on awareness of AI in healthcare, trust in AI systems, perceived efficiency of AI, preferences for personal interaction, and overall receptivity of AI in DR screening. Responses were stratified by sociodemographic and neighborhood characteristics.Results: A total of 100 participants were included (mean age 60 years; 52% female; 24% Hispanic, 20% non-Hispanic Black, 31% non-Hispanic White, 25% other). Most were aware of AI (78%) and its use in healthcare (70%), but fewer of its application to eye disease (46%). Most believed AI could improve accuracy and protect confidentiality (both 77%), yet 83% preferred physician oversight and would trust AI more if it were supervised by a doctor. Overall, 76% were comfortable with AI as part of the eye exam and 92% were satisfied with AI-based screening. Despite this, only 31% felt AI could replace a doctor visit and the majority of participants (94%) believed doctors will always remain responsible for diagnosing, even if AI was evaluating the scans.Conclusion: We found that while participants were generally comfortable with the use of AI as part of the eye examination, most would trust AI more if it were supervised by a doctor and did not believe that AI-based DR screening could replace a doctor’s visit. Implementation of autonomous AI-based DR screening should address the mismatch between the intended use of fully autonomous AI screening and participants’ understanding of the role that would mean for physicians.Keywords: artificial intelligence, diabetic retinopathy, fundus photography, health disparities