Benchmarking Differentially Private Residual Networks for Medical Imagery
This work addresses privacy concerns in medical imaging for healthcare applications, but it is incremental as it benchmarks existing methods.
The paper measured the effectiveness of ε-Differential Privacy (DP) in medical imaging by comparing Local-DP and DP-SGD, analyzing trade-offs between model accuracy and privacy guarantees, and evaluating their real-world utility.
In this paper we measure the effectiveness of $ε$-Differential Privacy (DP) when applied to medical imaging. We compare two robust differential privacy mechanisms: Local-DP and DP-SGD and benchmark their performance when analyzing medical imagery records. We analyze the trade-off between the model's accuracy and the level of privacy it guarantees, and also take a closer look to evaluate how useful these theoretical privacy guarantees actually prove to be in the real world medical setting.