Artificial intelligence versus manual screening for the detection of diabetic retinopathy: a comparative systematic review and meta-analysis
Bibliographic record
Abstract
Background: Diabetic retinopathy is one of the leading causes of blindness globally, among individuals with diabetes mellitus. Early detection through screening can help in preventing disease progression. In recent advancements artificial Intelligence assisted screening has emerged as an alternative to traditional manual screening methods. This diagnostic test accuracy (DTA) review aims to compare the sensitivity and specificity of AI versus manual screening for detecting diabetic retinopathy, focusing on both dilated and un-dilated eyes. Methods: A systematic review and meta-analysis were conducted for comparison of AI vs. manual screening of diabetic retinopathy using 25 observational (cross sectional, validation and cohort) studies with total images of 613,690 used for screening published between January 2015 and December 2024. Outcomes of the study was sensitivity, and specificity. Risk of bias was assessed using the QUADAS-2 tool for validation studies, the AXIS tool for cross-sectional studies, and the Newcastle-Ottawa Scale for cohort studies. Results: The results of this meta-analysis showed that for un-dilated eyes, AI screening showed pooled sensitivity of 0.90 [95% CI: 0.85-0.94] and pooled specificity of 0.94 [95% CI: 0.91-0.96] while manual screening shows pooled sensitivity of 0.79 [95% CI: 0.60-0.91] and pooled specificity of 0.99 [95% CI: 0.98-0.99]. For dilated eyes the pooled sensitivity of AI screening is 0.95 [95% CI: 0.91-0.97] and pooled specificity is 0.87 [95% CI: 0.79-0.92], while manual screening sensitivity is 0.90 [95% CI: 0.87-0.92] and specificity is 0.99 [95% CI: 0.99-1.00]. These data show comparable sensitivities and specificities of AI and manual screening, with AI performing better in sensitivity. Conclusion: AI-assisted screening for diabetic retinopathy shows comparable sensitivity and specificity compared to manual screening. These results suggest that AI can be a reliable alternative in clinical settings, with increased early detection rates and reducing the burden on ophthalmologists. Further research is needed to validate these findings. Systematic review registration: https://www.crd.york.ac.uk/PROSPERO/home, CRD42024596611.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.012 | 0.002 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".