Retinal Detachment
Conditions
Interventions
None listed
Sponsors
Yokohama City University
Eligibility
Sex/Gender
All
Inclusion criteria
Inclusion criteria: This meta-analysis includes studies that evaluate the diagnostic accuracy - specifically sensitivity and specificity - of deep learning algorithms using ultra-widefield fundus images for the detection of retinal detachment. Eligible study designs include case-control, cross-sectional, prospective, and retrospective observational studies. Studies must report sufficient data to allow calculation of diagnostic accuracy metrics (true positives, false positives, false negatives, true negatives). While peer-reviewed original articles are primarily targeted, conference abstracts may also be included if they contain adequate diagnostic information.
Exclusion criteria
Exclusion criteria: Studies that report only sensitivity or specificity, or that lack sufficient data to calculate diagnostic accuracy metrics, will be excluded. In addition, studies suspected of using duplicate datasets and non-original articles such as case reports, reviews, and editorials will also be excluded.
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| Sensitivity and specificity of deep learning algorithms using ultra-widefield fundus images for the diagnosis of retinal detachment, based on the assessment at the time of diagnosis in each included study | — |
Secondary
| Measure | Time frame |
|---|---|
| Key secondary outcomes include the area under the summary receiver operating characteristic curve (AUC) at the time of diagnosis, the diagnostic odds ratio (DOR), between-study heterogeneity assessed by the I^2 statistic, and risk of bias evaluated using the QUADAS-2 tool. | — |
Countries
Japan
Contacts
Public ContactYuki Mizuki
Yokohama City University Hospital Department of Ophthalmology
Outcome results
None listed