Changes in Diabetic Retinopathy Severity Levels Identified from Retinal Images Interpretation by Deep Learning Algorithms, Trained Graders, and Retinal Specialists.
- Conditions
- Diabetic retinopathy
- Registration Number
- TCTR20180716003
- Lead Sponsor
- /A
- Brief Summary
Not available
- Detailed Description
Not available
Recruitment & Eligibility
- Status
- Active, not recruiting
- Sex
- All
- Target Recruitment
- 7450
- Patients with diabetes who had no DR or mild NPDR and had their retinal images taken in the National DR Screening Program in Ophthalmic Service Plan of Ministry of Public Health of Thailand during the fiscal years of 2015 to 2017
- The patients described above had to have at least two retinal images taken in both the years 2015 and 2017.
- Diabetic patients whose retinal images cannot be graded for DR will be counted as referrals by other criteria but not be counted for estimation of the incidence of referrals of DR
- Diabetic patients whose one of the required retinal images cannot be transferred completely into the central server due to whatever reasons.
- Diabetic patients with cornea, lens, or other retinal abnormalities such as trauma, glaucoma, cataract, uveitis, vitreous hemorrhage from other causes precluded grading of DR
Study & Design
- Study Type
- Observational
- Study Design
- Not specified
- Primary Outcome Measures
Name Time Method incidence of diabetic retinopathy 2 years Graders, Deep learning algorithm and Retinal specialist
- Secondary Outcome Measures
Name Time Method To compare the incidence of diabetic retinopathy with gold standard (reitnal specialist) 2 years Graders, Deep learning algorithm and Retinal specialist,To compare the incidence of diabetic retinopathy between graders and deep learning algorithm 2 years Graders and Deep learning algorithm