CTF-Net

CTF-Net: Retinal Vessel Segmentation via Deep Coarse-to-Fine Supervision Network

Requirement

• Python3.6
• PyTorch 0.4
• configparser

How to run(CTF—Net)

• run python prepare_datasets_DRIVE.py to generate hdf5 file of training data
• run cd src
• run python retinaNN_training.py to train
• run python retinaNN_predict.py to test

Parameter defination

• parameters (path, patch size, et al.) are defined in "configuration.txt"
• training parameters are defined in src/retinaNN_training.py line 49 t 84 with notes "=====Define parameters here ========="

Pretrained weights

• pretrained weights are stored in "src/checkpoint"
• results are stored in "test/"

Results

The results reported in the ./test folder are referred to the trained model which reported the minimum validation loss. The ./test folder includes:

• Model:
  • test_model.png schematic representation of the neural network
  • test_architecture.json description of the model in json format
  • test_best_weights.h5 weights of the model which reported the minimum validation loss, as HDF5 file
  • test_last_weights.h5 weights of the model at last epoch (150th), as HDF5 file
  • test_configuration.txt configuration of the parameters of the experiment
• Experiment results:
  • performances.txt summary of the test results, including the confusion matrix
  • Precision_recall.png the precision-recall plot and the corresponding Area Under the Curve (AUC)
  • ROC.png the Receiver Operating Characteristic (ROC) curve and the corresponding AUC
  • all_*.png the 20 images of the pre-processed originals, ground truth and predictions relative to the DRIVE testing dataset
  • sample_input_*.png sample of 40 patches of the pre-processed original training images and the corresponding ground truth
  • test_Original_GroundTruth_Prediction*.png from top to bottom, the original pre-processed image, the ground truth and the prediction. In the predicted image, each pixel shows the vessel predicted probability, no threshold is applied.

The following table compares this method to other recent techniques, which have published their performance in terms of Area Under the ROC curve (AUC ROC) on the DRIVE dataset.

Citation

• If you find this work is helpful, please cite our paper

@inproceedings{wang2020ctf,
  title={CTF-Net: Retinal Vessel Segmentation via Deep Coarse-To-Fine Supervision Network},
  author={Wang, Kun and Zhang, Xiaohong and Huang, Sheng and Wang, Qiuli and Chen, Feiyu},
  booktitle={2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI)},
  pages={1237--1241},
  year={2020},
  organization={IEEE}
}