REAN: Reconstruction-aware ECG Anonymization Based on Privacy--Utility Orthogonality

2026-07-07

AI Summary

This study addresses the challenge of balancing privacy preservation and diagnostic utility in electrocardiogram (ECG) sharing by proposing REAN, a reconstruction-aware anonymization method. REAN leverages a 1D U-Net to reconstruct ECG signals under the supervision of frozen privacy and utility classifiers, and uniquely enforces near-orthogonality between their gradients—approximately 93.8°—to achieve effective decoupling of privacy and utility. Evaluated on four PhysioNet datasets, REAN reduces re-identification rates to near-random levels (from 0.96 to 0.00) while preserving arrhythmia detection performance, achieving a macro AUROC of 0.9982 compared to 0.9991 on the original data. Furthermore, the method demonstrates robustness against unseen privacy classifier architectures.

Abstract

A shared electrocardiogram (ECG) is itself a biometric fingerprint that can re-identify a patient and reveal personal information. Recent ECG anonymizers transform the signal before sharing to reduce privacy leakage. However, existing methods still face a privacy--utility trade-off, in which preserving privacy often compromises utility while preserving utility reveals personal information. We propose \emph{REAN} (\emph{RE}construction-aware ECG \emph{AN}onymizer), a raw ECG signal anonymizer, to address this privacy--utility trade-off. REAN reconstructs the signal using a 1-D U-Net trained with losses from frozen privacy and utility classifiers to reduce privacy leakage while preserving utility. The privacy and utility gradients are near-orthogonal ($\approx$93.8$^\circ$), so reducing privacy leakage leaves utility almost unchanged. On four public PhysioNet databases, REAN achieves the strongest privacy--utility balance among raw ECG signal baselines. It drives re-identification to chance (0.96$\to$0.00), keeps arrhythmia macro-AUROC at the clean level (Clean 0.9982 vs.\ REAN 0.9991), and maintains re-identification protection under unseen privacy-classifier architectures.

Problem

• ECG anonymization
• privacy--utility trade-off
• biometric re-identification
• signal utility
• privacy leakage

Innovation

• ECG anonymization
• privacy–utility trade-off
• gradient orthogonality
• reconstruction-aware learning
• biometric privacy