Related Experiment Video
Updated: May 6, 2026
09:17
Using Retinal Imaging to Study Dementia
Published on: November 6, 2017
21.6K
Accuracy of Autonomous Artificial Intelligence-Based Diabetic Retinopathy Screening in Real-Life Clinical Practice
Eleonora Riotto1, Stefan Gasser1, Jelena Potic1
1Hôpital Jules Gonin, 1004 Lausanne, Switzerland.
Journal of Clinical Medicine
|August 29, 2024
Summary
The IDX-DR AI system for diabetic retinopathy (DR) shows high accuracy, especially for negative results. However, it tends to overestimate DR severity, requiring specialist review for mild or greater classifications.
Area of Science:
- Ophthalmology
- Artificial Intelligence
- Medical Imaging
Background:
- Diabetic retinopathy (DR) detection is crucial for preventing vision loss.
- Artificial intelligence (AI) and machine learning offer promising new diagnostic tools.
- The IDX-DR machine utilizes AI and deep learning for DR identification.
Purpose of the Study:
- To evaluate the diagnostic performance of the IDX-DR machine.
- To compare AI-based DR detection with expert human grading.
- To assess the accuracy, sensitivity, and specificity of the IDX-DR system.
Main Methods:
- Retrospective review of 2282 retinal images from 1141 patients.
- Comparison of IDX-DR machine classifications with two experienced retinal specialists' gradings.
- Calculation of sensitivity, specificity, PPV, NPV, and accuracy.
Main Results:
- IDX-DR achieved 100% sensitivity for 'no DR', 'mild DR', and 'moderate DR'.
- Specificity ranged from 78.4% ('no DR') to 97.6% ('severe DR').
- Negative predictive value (NPV) was 100% across all categories; accuracy exceeded 80% for early stages.
Conclusions:
- The IDX-DR machine demonstrates high reliability for ruling out diabetic retinopathy.
- The system tends to overestimate DR severity, necessitating specialist confirmation for positive findings.
- Clinicians can trust negative results from the IDX-DR screening software.
