Detecting Center-Involved Diabetic Macular Edema from Analysis of Retina Images Using Deep Learning
Not Applicable
Recruiting
- Conditions
- Diabetic macular edemaDiabetic retinopathy,Diabetic macular edema,Accuracy,Artificial intelligence,Deep-learning,Artificial neural network,Sensitivity,Specificity
- Registration Number
- TCTR20180818002
- Lead Sponsor
- Rajavithi hospital
- Brief Summary
Not available
- Detailed Description
Not available
Recruitment & Eligibility
- Status
- Recruiting
- Sex
- All
- Target Recruitment
- 7500
Inclusion Criteria
Diabetic patient with retinal images(45 degree with macular-center) and OCT images which is macular dense volume scan (49 high-speed B-scans [512 A scan/B scan]; 20x20 degree, 6x6 mm were taken on the same date.
Naiive patients with no previous treatment
Exclusion Criteria
Patient with retinal or macular disease which distort the interpretation of OCT images such as choroidal neovascularization, retinal vein occlusion, postsurgical macular edema, central serous chorioretinopathy, macular retinal detachment, epiretinal membrane, macular hole, or vitreomacular traction
retinal images with poor quality
Study & Design
- Study Type
- Observational
- Study Design
- Not specified
- Primary Outcome Measures
Name Time Method To develope the deep learning system from OCT imaging to detect diabetic macular edema from color fu 1 year Algorithm
- Secondary Outcome Measures
Name Time Method Effectiveness of deep learning algorithm in screening for diabetic macular edema 1 year sensitivity,specificity, accuracy, area under the curve comparing with retinal specialists