USQNet

This is the official source code repository for the paper titled "Deep Learning Model for Quality Assessment of Urinary Bladder Ultrasound Images using Multi-scale and Higher-order Processing" (Paper link). The focus of our work is to address the challenges in Autonomous Ultrasound Image Quality Assessment (US-IQA). Plase refer to the paper for more details.

Installations

conda env create -f environment.yml

Training

python train.py --data_dir dataset/{dir_name} --num_epochs --folds --lr --batch_size

The trained model will be saved in the outputs folder.

Testing

python test.py --data_dir dataset/{dir_name} --load_model outputs/{model_name}.pth

Citation

(will be updated soon)

@article{raina2024deep,
  title={Deep Learning Model for Quality Assessment of Urinary Bladder Ultrasound Images using Multi-scale and Higher-order Processing},
  author={Raina, Deepak and SH, Chandrashekhara and Voyles, Richard and Wachs, Juan and Saha, Subir Kumar},
  journal={IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control},
  volume={??},
  number={?},
  pages={???--???},
  year={2024},
  publisher={IEEE}
}

License

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License for Noncommercial use only. Any commercial use should obtain formal permission.

Acknowledgement

This code base is built upon ResNet, Res2Net and MPN-COV. Thanks to the authors of these papers for making their code available for public usage.