English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Privacy regulations and the physical distribution of heterogeneous data areoften primary concerns for the development of deep learning models in a medicalcontext. This paper evaluates the feasibility of differentially privatefederated learning for chest X-ray classification as a defense against dataprivacy attacks. To the best of our knowledge, we are the first to directlycompare the impact of differentially private training on two different neuralnetwork architectures, DenseNet121 and ResNet50. Extending the federatedlearning environments previously analyzed in terms of privacy, we simulated aheterogeneous and imbalanced federated setting by distributing images from thepublic CheXpert and Mendeley chest X-ray datasets unevenly among 36 clients.Both non-private baseline models achieved an area under the receiver operatingcharacteristic curve (AUC) of $0.94$ on the binary classification task ofdetecting the presence of a medical finding. We demonstrate that both modelarchitectures are vulnerable to privacy violation by applying imagereconstruction attacks to local model updates from individual clients. Theattack was particularly successful during later training stages. To mitigatethe risk of privacy breach, we integrated R\'enyi differential privacy with aGaussian noise mechanism into local model training. We evaluate modelperformance and attack vulnerability for privacy budgets $\epsilon \in$ {1, 3,6, 10}. The DenseNet121 achieved the best utility-privacy trade-off with an AUCof $0.94$ for $\epsilon$ = 6. Model performance deteriorated slightly forindividual clients compared to the non-private baseline. The ResNet50 onlyreached an AUC of $0.76$ in the same privacy setting. Its performance wasinferior to that of the DenseNet121 for all considered privacy constraints,suggesting that the DenseNet121 architecture is more robust to differentiallyprivate training.

Defending against Reconstruction Attacks through Differentially Private Federated Learning for Classification of Heterogeneous Chest X-Ray Data