ViTReg-IP Lung Analysis Architecture

What is ViTReg-IP?

ViTReg-IP combines feature extraction models and is based on a vision transformer (ViT) that acts as a backbone for a regression (Reg) head targeting infection prediction (IP) of lungs from chest radiographs. This approach allows precise and automatic quantification of the severity of lung infection. Experimental results have shown that the proposed model has the ability and efficiency to provide peak performance in quantifying severity with high generalizability. This architecture is developed in the context of a project ViTAL which aims to develop Vision-based Techniques for Analyzing lungs and to continously propose solutions through this open-source initiative. http://dx.doi.org/10.1007/s11517-024-03066-3

Team

Core Contributors

• Biomedical Engineer, PhD Student, Bouthaina Slika, University of the Basque Country, Spain, [email protected]
• Prof. Dr. Fadi Dornaika, IEEE member, Dept. of Artificial Intelligence, University of the Basque Country & IKERBAQUE foundation, Spain [email protected]
• Prof. Dr. Karim Hammoudi, IEEE member, Group Imagery, Dept. of Computer Science, IRIMAS, Université de Haute-Alsace, France, [email protected]
• Medical Doctor, PhD, Hamid Merdji, French National Institute of Health and Medical Research (INSERM), Regenerative Nanomedicine (RNM), Biomedicine Research Center (CRBS), Federation of Translational Medicine, and Dept. of Intensive Medicine-Resuscitation, Hospital of Strasbourg, France, [email protected]
• Dr. Vinh Truong Hoang, Dept. of Computer Science, HCMC Open University, Ho Chi Minh City, Vietnam, [email protected]
• Iyed Dhahri, Final Year Engineering Student, ENSI, Tunisia

Collaborators

• Prof. Dr. Halim Benhabiles, Group BIO-MEMS, Dept. of Artificial Intelligence, JUNIA, CNRS, IEMN, University of Lille, [email protected]
• Prof. Dr. Mahmoud Melkemi, Group Imagery, Dept. of Computer Science, IRIMAS, Université de Haute-Alsace, [email protected]
• Prof. Dr. Adnance Cabani, Dept. of Computer Science, ESIGELEC/IRSEEM, Normandy University, [email protected]

Reference

Karim Hammoudi, Halim Benhabiles, Mahmoud Melkemi, Fadi Dornaika, Ignacio Arganda-Carreras, Dominique Collard, and Arnaud Scherpereel. Deep learning on chest x-ray images to detect and evaluate pneumonia cases at the era of covid-19. Journal of medical systems, 45(7):1–10, 2021. https://doi.org/10.1007/s10916-021-01745-4

@Article{Hammoudi2021,
author={Hammoudi, Karim
and Benhabiles, Halim
and Melkemi, Mahmoud
and Dornaika, Fadi
and Arganda-Carreras, Ignacio
and Collard, Dominique
and Scherpereel, Arnaud},
title={Deep Learning on Chest X-ray Images to Detect and Evaluate Pneumonia Cases at the Era of COVID-19},
journal={Journal of Medical Systems},
year={2021},
month={June},
day={08},
issn={1573-689X},
doi={10.1007/s10916-021-01745-4},
url={https://doi.org/10.1007/s10916-021-01745-4}
}

Bouthaina Slika, Fadi Dornaika, Karim Hammoudi, and Vinh Truong Hoang. Automatic quantification of lung infection severity in chest x-ray images. In IEEE Statistical Signal Processing (SSP) Workshop, pages 418–422. IEEE, 2023.

@inproceedings{slika2023ssp,
title={Automatic Quantification of Lung Infection Severity in Chest X-ray Images},
author={Slika, Bouthaina
and Dornaika, Fadi
and Hammoudi, Karim
and Hoang, Vinh Truong},
booktitle={IEEE Statistical Signal Processing (SSP) Workshop},
pages={418--422},
year={2023},
organization={IEEE}
}