Attention-Guided Discriminative Region Localization for Bone Age Assessment

• This repository contains code for our paper Attention-Guided Discriminative Region Localization for Bone Age Assessment Download Paper
• If you have any question about our paper or code, please don't hesitate to contact with me [email protected], we will update our repository accordingly

Setup

• Dataset The RSNA dataset can be downloaded here RSNA

• requirements Python==3.6, tensorflow==1.9, keras = 2.1.6, opencv.

Training

• Step 1: Generate *.npy Data with data_utils.py
• Step 2: Run main_classification.py to train classification model, the Attention Map and Heatmap will be saved according to the given path.
• The attention map will be generated by GAPAttention(model,weights,'/raid/chenchao/code/BoneAge/BoneAge/data/train/')
• Step 3: Generate the Heatmap for the Hand Reigon, Region-1 and Region-2 one-by-one
• Step 4: Run data/crop_patches.py to crop the local patches for Hand, Region-1 and Region-2 according to the heatmap one-by-one.
• Step 5: Run main_aggregation.py to aggragate different local patches for BAA
• You can also run main_regression.py to get the BAA performance by using one local patch.

Results

Citation

• If you find it helpful for you, please cite our paper

@article{chen2020attention,
  title={Attention-Guided Discriminative Region Localization for Bone Age Assessment},
  author={Chen, Chao and Chen, Zhihong and Jin, Xinyu and Li, Lanjuan and Speier, William and Arnold, Corey W},
  journal={arXiv preprint arXiv:2006.00202},
  year={2020}
}