CXR-FL: Deep Learning-Based Chest X-ray Image Analysis Using Federated Learning

Filip Ślazyk; Przemysław Jabłecki; Aneta Lisowska; Maciej Malawski; Szymon Płotka

doi:10.1007/978-3-031-08754-7_50

CXR-FL: Deep Learning-Based Chest X-ray Image Analysis Using Federated Learning

Filip Ślazyk, Przemysław Jabłecki, Aneta Lisowska, Maciej Malawski, Szymon Płotka

Research output: Chapter in Book / Report / Conference proceeding › Conference contribution › Academic › peer-review

Abstract

Federated learning enables building a shared model from multicentre data while storing the training data locally for privacy. In this paper, we present an evaluation (called CXR-FL) of deep learning-based models for chest X-ray image analysis using the federated learning method. We examine the impact of federated learning parameters on the performance of central models. Additionally, we show that classification models perform worse if trained on a region of interest reduced to segmentation of the lung compared to the full image. However, focusing training of the classification model on the lung area may result in improved pathology interpretability during inference. We also find that federated learning helps maintain model generalizability. The pre-trained weights and code are publicly available at (https://github.com/SanoScience/CXR-FL ).

Original language	English
Title of host publication	Computational Science - ICCS 2022, 22nd International Conference, Proceedings
Editors	D. Groen, M. Paszynski, J.J. Dongarra
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	433-440
ISBN (Print)	9783031087530
DOIs	https://doi.org/10.1007/978-3-031-08754-7_50
Publication status	Published - 2022
Externally published	Yes
Event	22nd Annual International Conference on Computational Science, ICCS 2022 - London, United Kingdom Duration: 21 Jun 2022 → 23 Jun 2022

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	22nd Annual International Conference on Computational Science, ICCS 2022
Country/Territory	United Kingdom
City	London
Period	21/06/22 → 23/06/22

Funding

Acknowledgements. This publication is partly supported by the EU H2020 grant Sano (No. 857533) and the IRAP Plus programme of the Foundation for Polish Science. This research was supported in part by the PL-Grid Infrastructure. We would like to thank Piotr Nowakowski for his assistance with proofreading the manuscript.

Funders	Funder number
EU H2020	857533
Fundacja na rzecz Nauki Polskiej

Access to Document

10.1007/978-3-031-08754-7_50

Persistent URL (handle)

Copy link address

Cite this

Ślazyk, F., Jabłecki, P., Lisowska, A., Malawski, M., & Płotka, S. (2022). CXR-FL: Deep Learning-Based Chest X-ray Image Analysis Using Federated Learning. In D. Groen, M. Paszynski, & J. J. Dongarra (Eds.), Computational Science - ICCS 2022, 22nd International Conference, Proceedings (pp. 433-440). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-08754-7_50

Ślazyk, Filip ; Jabłecki, Przemysław ; Lisowska, Aneta et al. / CXR-FL : Deep Learning-Based Chest X-ray Image Analysis Using Federated Learning. Computational Science - ICCS 2022, 22nd International Conference, Proceedings. editor / D. Groen ; M. Paszynski ; J.J. Dongarra. Springer Science and Business Media Deutschland GmbH, 2022. pp. 433-440 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{c5be7f884cce419a90c730930936210f,

title = "CXR-FL: Deep Learning-Based Chest X-ray Image Analysis Using Federated Learning",

abstract = "Federated learning enables building a shared model from multicentre data while storing the training data locally for privacy. In this paper, we present an evaluation (called CXR-FL) of deep learning-based models for chest X-ray image analysis using the federated learning method. We examine the impact of federated learning parameters on the performance of central models. Additionally, we show that classification models perform worse if trained on a region of interest reduced to segmentation of the lung compared to the full image. However, focusing training of the classification model on the lung area may result in improved pathology interpretability during inference. We also find that federated learning helps maintain model generalizability. The pre-trained weights and code are publicly available at (https://github.com/SanoScience/CXR-FL ).",

author = "Filip {\'S}lazyk and Przemys{\l}aw Jab{\l}ecki and Aneta Lisowska and Maciej Malawski and Szymon P{\l}otka",

year = "2022",

doi = "10.1007/978-3-031-08754-7_50",

language = "English",

isbn = "9783031087530",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "433--440",

editor = "D. Groen and M. Paszynski and J.J. Dongarra",

booktitle = "Computational Science - ICCS 2022, 22nd International Conference, Proceedings",

note = "22nd Annual International Conference on Computational Science, ICCS 2022 ; Conference date: 21-06-2022 Through 23-06-2022",

}

Ślazyk, F, Jabłecki, P, Lisowska, A, Malawski, M & Płotka, S 2022, CXR-FL: Deep Learning-Based Chest X-ray Image Analysis Using Federated Learning. in D Groen, M Paszynski & JJ Dongarra (eds), Computational Science - ICCS 2022, 22nd International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Springer Science and Business Media Deutschland GmbH, pp. 433-440, 22nd Annual International Conference on Computational Science, ICCS 2022, London, United Kingdom, 21/06/22. https://doi.org/10.1007/978-3-031-08754-7_50

CXR-FL: Deep Learning-Based Chest X-ray Image Analysis Using Federated Learning. / Ślazyk, Filip; Jabłecki, Przemysław; Lisowska, Aneta et al.
Computational Science - ICCS 2022, 22nd International Conference, Proceedings. ed. / D. Groen; M. Paszynski; J.J. Dongarra. Springer Science and Business Media Deutschland GmbH, 2022. p. 433-440 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

Research output: Chapter in Book / Report / Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - CXR-FL

T2 - 22nd Annual International Conference on Computational Science, ICCS 2022

AU - Ślazyk, Filip

AU - Jabłecki, Przemysław

AU - Lisowska, Aneta

AU - Malawski, Maciej

AU - Płotka, Szymon

PY - 2022

Y1 - 2022

N2 - Federated learning enables building a shared model from multicentre data while storing the training data locally for privacy. In this paper, we present an evaluation (called CXR-FL) of deep learning-based models for chest X-ray image analysis using the federated learning method. We examine the impact of federated learning parameters on the performance of central models. Additionally, we show that classification models perform worse if trained on a region of interest reduced to segmentation of the lung compared to the full image. However, focusing training of the classification model on the lung area may result in improved pathology interpretability during inference. We also find that federated learning helps maintain model generalizability. The pre-trained weights and code are publicly available at (https://github.com/SanoScience/CXR-FL ).

AB - Federated learning enables building a shared model from multicentre data while storing the training data locally for privacy. In this paper, we present an evaluation (called CXR-FL) of deep learning-based models for chest X-ray image analysis using the federated learning method. We examine the impact of federated learning parameters on the performance of central models. Additionally, we show that classification models perform worse if trained on a region of interest reduced to segmentation of the lung compared to the full image. However, focusing training of the classification model on the lung area may result in improved pathology interpretability during inference. We also find that federated learning helps maintain model generalizability. The pre-trained weights and code are publicly available at (https://github.com/SanoScience/CXR-FL ).

UR - http://www.scopus.com/inward/record.url?scp=85134349892&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-08754-7_50

DO - 10.1007/978-3-031-08754-7_50

M3 - Conference contribution

SN - 9783031087530

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 433

EP - 440

BT - Computational Science - ICCS 2022, 22nd International Conference, Proceedings

A2 - Groen, D.

A2 - Paszynski, M.

A2 - Dongarra, J.J.

PB - Springer Science and Business Media Deutschland GmbH

Y2 - 21 June 2022 through 23 June 2022

ER -

Ślazyk F, Jabłecki P, Lisowska A, Malawski M, Płotka S. CXR-FL: Deep Learning-Based Chest X-ray Image Analysis Using Federated Learning. In Groen D, Paszynski M, Dongarra JJ, editors, Computational Science - ICCS 2022, 22nd International Conference, Proceedings. Springer Science and Business Media Deutschland GmbH. 2022. p. 433-440. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-08754-7_50

CXR-FL: Deep Learning-Based Chest X-ray Image Analysis Using Federated Learning

Abstract

Publication series

Conference

Funding

Access to Document

Persistent URL (handle)

Other files and links

Fingerprint

Cite this