dr-models

Diabetic Retinopathy Classifiers

Model for detecting Diabetic Retinopathy from retinal fundus images.

Trained as part of the paper: Automated and Explainable Detection of Multiple Diseases from Retinal Fundus Images

Model Architectures

Loading any one model

from keras.models import model_from_json
import json

model = keras.Model.from_config(config)
model.load_weights("model.weights.h5")

with open("deep_learning/ResNet50_enhanced/ResNet50_pretrained_2000_enhanced.json", 'r') as json_file:
    model_json = json_file.read()
model = model_from_json(model_json)
model.load_weights("deep_learning/ResNet50_enhanced/ResNet50_pretrained_2000_enhanced.weights.h5")
model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])

Best Performance (Enhanced ResNet50 backbone)

Presence

• Accuracy: 93.4%
• Precision: 92.34%
• Recall: 94.6%

Grading

• Accuracy: 79.35%
• Precision: 79.15%
• Recall: 79.35%
• Cohen's Kappa (linear): 82.89%
• Cohen's Kappa (quadratic): 90.67%

Citation

@inproceedings{masti2026automated,
  title={Automated and Explainable Detection of Multiple Diseases from Retinal Fundus Images},
  author={Masti, Shubha and Prasad, T. and Srinivasa, G.},
  booktitle={Image Processing and Vision Engineering. IMPROVE 2025},
  series={Communications in Computer and Information Science},
  volume={2628},
  publisher={Springer},
  year={2026},
  doi={10.1007/978-3-032-01169-5_9}
}