OCT Subretinal Fluid Classifier — CSC / PCV / VKH

Automated classification of subretinal fluid (SRF) accumulation from sequences of retinal optical coherence tomography (OCT) scans into three conditions:

Label	Full name
CSC	Central Serous Chorioretinopathy
PCV	Polypoidal Choroidal Vasculopathy
VKH	Vogt-Koyanagi-Harada disease
NoSRF	Normal

This model accompanies the peer-reviewed publications:

"Diagnostic Performance of Machine Learning Technology Using Optical Coherence Tomographic Image in Retinal Diseases Presented with Subretinal Fluid" Khakhai, Panita and Paisarnsrisomsuk, Sarun and Wiratchotisatian, Pitchaya and Tangprasert, Kwankhao and Laovirojjanakul, Wipada — Retina, 2026 DOI: 10.1097/IAE.0000000000004708

"Enhanced Retinal-Choroidal Disorders Classification Model via Temporal Sequence Analysis of OCT Images Across Multiple Lines of Fovea" Tangprasert, Kwankhao and Wiratchotisatian, Pitchaya and Khakhai, Panita and Laovirojjanakul, Wipada and Paisarnsrisomsuk, Sarun — 2025 ICAIIC, IEEE, 2025 DOI: 10.1109/ICAIIC64266.2025.10920750

📦 Code, notebooks, and full documentation: GitHub — oct-srf-classifier

Intended Use

✅ Appropriate uses

Academic research on retinal disease classification from OCT
Benchmarking against other automated OCT analysis methods
Educational demonstration of Grad-CAM explainability in ophthalmology

❌ Not appropriate for

Standalone clinical diagnosis without physician review
Deployment in patient-facing systems without local validation and regulatory approval
Use on patient populations significantly different from the Thai clinical cohort

Limitations and Bias

Single-centre, single-country data. The training dataset comes exclusively from one Thai referral hospital. Performance on other ethnic populations or devices from different manufacturers has not been evaluated.
No concurrent conditions. Cases with co-existing retinal pathologies were not included during training.

Ethical Statement

Training data was collected under Institutional Review Board (IRB) approval at Khon Kaen University Hospital, Thailand. All patient data is de-identified and is not publicly released due to privacy regulations and hospital ethics policy. Only model weights are distributed through this repository.

Researchers wishing to inquire about dataset access for independent verification should contact the corresponding author listed in the paper.

Citation

If you use this model or code in your research, please cite the original paper:


@article{khakhai2026diagnostic,
  title   = {Diagnostic Performance of Machine Learning Technology Using Optical Coherence Tomographic Image in Retinal Diseases Presented with Subretinal Fluid},
  author  = {Khakhai, Panita and Paisarnsrisomsuk, Sarun and Wiratchotisatian, Pitchaya and Tangprasert, Kwankhao and Laovirojjanakul, Wipada},
  journal = {Retina},
  volume  = {46},
  number  = {3},
  pages   = {514--520},
  year    = {2026},
  month   = {March},
  doi     = {10.1097/IAE.0000000000004708}
}

@INPROCEEDINGS{10920750,
  author={Tangprasert, Kwankhao and Wiratchotisatian, Pitchaya and Khakhai, Panita and Laovirojjanakul, Wipada and Paisarnsrisomsuk, Sarun},
  booktitle={2025 International Conference on Artificial Intelligence in Information and Communication (ICAIIC)}, 
  title={Enhanced Retinal-Choroidal Disorders Classification Model via Temporal Sequence Analysis of OCT Images Across Multiple Lines of Fovea}, 
  year={2025},
  volume={},
  number={},
  pages={0766-0771},
  keywords={Analytical models;Accuracy;Fluids;Sequences;Decision making;Predictive models;Streaming media;Retina;Real-time systems;Diseases;Deep Learning;LSTM;CNN;Medical Image Analysis;Retinal OCT Image;Subretinal Fluid Classification;Central Serous Chorioretinopathy;Polypoidal Choroidal Vasculopathy;Vogt-Koyanagi-Harada},
  doi={10.1109/ICAIIC64266.2025.10920750}
}

Contact

For questions about this model, please open an issue on the GitHub repository or contact the corresponding author listed in the IEEE publication. Khon Kaen University — Department of Computer Engineering, Thailand hon Kaen University- Department of Statistic, Thailand Khon Kaen University — Faculty of Medicine, Department of Ophthalmology, Srinagarind Hospital, Thailand

Downloads last month: -; Downloads are not tracked for this model. How to track