Automated Explainable Multidimensional Deep Learning Platform of Retinal Images for Retinopathy of Prematurity Screening
importance 6/10
ai assisted
Retinopathy of Prematurity Studies neonatal ophthalmology stage: research
Problem
ROP diagnosis relies on indirect ophthalmoscopy by experienced ophthalmologists, which may limit early detection and timely treatment.
Why it matters
Early detection and treatment of retinopathy of prematurity improves visual outcomes in preterm infants.
What the AI did
A multidimensional deep learning platform with four classifiers (image quality, any ROP stage, intraocular hemorrhage, preplus/plus disease) automatically generates referral-warranted ROP assessments.
Results
not stated in source
Impact
Developed and validated on 52,249 retinal images from 8,652 preterm infants across 4 centers.
Record
| Venue | JAMA Network Open |
|---|---|
| Published | 2021 |
| Work type | article |
| Authors | Ji Wang, Jie Ji, Mingzhi Zhang, Jianwei Lin, Guihua Zhang, Weifen Gong, Ling‐Ping Cen, Yamei Lu et al. |
| Citations | 83 (OpenAlex, at import time) |
| References | 37 |
| Model or method | not stated in source |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1001/jamanetworkopen.2021.8758 |
| Open access | gold |
| Keywords | retinopathy of prematurity, deep learning, retinal images, medical screening, explainable AI |
Importance rating
6
significant — thousands to millions affected
Automated ROP screening could enable early detection and treatment in preterm infants where expert ophthalmologist availability may be limited.