Few‑shot learning for COVID‑19 chest X‑ray classification with imbalanced data: an inter vs. intra domain study
Full text available here.
About • How To Use • Citations • License
Inter and Intra-domain study in few-shot learning scenarios with severe data imbalance based on Siamese neural networks. Tested over four chest X-ray datasets with annotated cases of both positive and negative COVID-19 diagnoses. All datasets are publicly accessible: ChestX-ray can be found at https://nihcc.app.box.com/v/ChestXray-NIHCC, GitHub-COVID at https://github.com/ieee8023/covid-chestxray-dataset, PadChest is available at https://bimcv.cipf.es/bimcv-projects/padchest, and BIMCV-COVID repositories can be accessed through https://bimcv.cipf.es/bimcv-projects/bimcv-covid19.
To replicate the work, execute the file main_launch_experiments.py. It is ready to receive different parameters, each one corresponding to a concrete experiment.
The code has been used over a Docker environment. However, the requirements for any other virtual environment can be easily extracted from docker/Dockerfile.
@Article{Galan-Cuenca2024,
author={Galan-Cuenca, Alejandro
and Gallego, Antonio Javier
and Saval-Calvo, Marcelo
and Pertusa, Antonio},
title={Few-shot learning for COVID-19 chest X-ray classification with imbalanced data: an inter vs. intra domain study},
journal={Pattern Analysis and Applications},
year={2024},
month={Jun},
day={11},
volume={27},
number={3},
pages={69},
issn={1433-755X},
doi={10.1007/s10044-024-01285-w},
url={https://doi.org/10.1007/s10044-024-01285-w}
}
This work is under a MIT license.