Hybrid deep ensemble architecture for robust diabetic retinopathy classification: leveraging transfer learning and CNN-transformer synergy
Youssef Maaod, Tamer Abuhmed, Eslam Amer, Shaker El-Sappagh · 2026
Read Hybrid deep ensemble architecture for robust diabetic retinopathy classification: leveraging transfer learning and CNN-transformer synergy by Youssef Maaod, Tamer Abuhmed, Eslam Amer, Shaker El-Sappagh (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13493815.txt.
Description
Maaod and colleagues built an artificial intelligence system to grade diabetic retinopathy, eye damage caused by diabetes that can blind people who miss regular screening. Their 2026 framework blends several model families: networks that read local image detail, vision transformers that weigh the whole picture, and transfer learning from pre-trained models. They screened 23 ready-made models on the APTOS 2019 dataset of retina photographs, then combined the strongest performers.
The winning design stacked the best hybrid models in sequence and merged their judgements by soft voting, where each model's confidence counts rather than a simple majority. That ensemble graded test images with accuracy in the low nineties and separated disease grades almost perfectly on the ranking measure reported. Combining different model types proved more stable than any single network. The authors pitch the system as a step toward dependable automated screening for clinics with few specialists.
Catalog record
| CREATOR | Youssef Maaod, Tamer Abuhmed, Eslam Amer, Shaker El-Sappagh |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 76 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |