JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Accuracy and safety of an autonomous artificial intelligence clinical assistant conducting telemedicine follow-up assessment for cataract surgery

  • 0Translational and Clinical Research Institute, Newcastle University, Newcastle upon Tyne, UK.
Eclinicalmedicine +

|

July 25, 2024

|

July 25, 2024

View abstract on PubMed

Summary

This summary is machine-generated.

An artificial intelligence (AI) system, Dora R1, accurately triages post-cataract surgery patients, showing high sensitivity and specificity. This AI tool offers a safe, cost-effective solution to expand clinical capacity and improve patient follow-up.

Area Of Science

  • Ophthalmology
  • Medical Artificial Intelligence
  • Digital Health

Background

  • Artificial intelligence (AI) can enhance patient triage post-cataract surgery, improving clinical capacity.
  • This study evaluates Dora R1, an autonomous telemedicine system, for post-cataract surgery patient assessment.
  • The AI's performance is compared against ophthalmic specialists.

Purpose Of The Study

  • To assess the accuracy and safety of the Dora R1 AI system in identifying cataract surgery patients requiring further management.
  • To compare Dora R1's triage decisions with those of human ophthalmic specialists.
  • To evaluate the usability, acceptability, and cost-effectiveness of the AI system.

Main Methods

  • 225 patients undergoing cataract surgery were recruited from UK teaching hospitals.
  • Dora R1 conducted autonomous follow-up calls, supervised in real-time by an ophthalmologist.
  • Primary analysis compared AI and clinician decisions; secondary analyses used mixed methods for usability, acceptability, and cost.

Main Results

  • Dora R1 achieved 94% sensitivity and 86% specificity, with strong agreement (kappa: 0.758-0.970) with clinicians.
  • The AI system demonstrated high feasibility (96.5% autonomous calls) and cost benefits (£35.18 per patient).
  • Patient acceptability was generally good, though concerns about the lack of human interaction in complex cases were noted.

Conclusions

  • The Dora R1 AI system shows preliminary evidence of safety, acceptability, feasibility, and cost-effectiveness for post-cataract surgery follow-up.
  • AI-driven telemedicine offers a viable approach to expand clinical capacity and support patient care.
  • Further real-world implementation studies are recommended to validate safety and effectiveness in diverse settings.

Keywords: