差分隐私与公平性审计的不规则纵向健康记录分数扩散
Differentially Private and Fairness-Audited Score Diffusion for Irregular Longitudinal Health Records
查看机构详情
- Superior University（苏佩里尔大学）
机构由 AI 辅助整理，请以论文原文为准。
浏览论文内容
中文总结 AI 辅助
提出TRUST LONGSYNTH，一种结合差分隐私与公平性审计的分数扩散生成器，用于不规则纵向健康记录，在隐私保护下提升合成数据效用并降低成员攻击风险。
中文摘要 AI 辅助
共享不规则纵向健康记录可以加速模型开发，然而合成数据的发布可能会泄露参与信息、扭曲时间依赖性、抑制罕见事件，或降低对代表性不足群体的效用。我们提出了TRUST LONGSYNTH，一个可审计的患者级隐私生成器，它结合了有界充分统计量、zCDP核算的高斯发布、条件解析分数扩散、块带状时间协方差、独立的缺失和间隙模型，以及带有总体权重的受保护事件采样下限。该方法在五个独立生成、三个队列基准上进行了评估，包含720名患者、十四个不规则观测时隙、六个混合变量、信息性缺失和罕见的恶化结果。在epsilon=12和delta=10的负5次方时，TRUST LONGSYNTH实现了平均训练合成测试真实AUPRC 0.342，Brier分数0.088，预期校准误差0.082，相关性误差0.222，自相关误差0.317，以及成员攻击AUROC 0.499。相对于私有对角分数基线，AUPRC提高了7.5%，而Brier、校准、相关性和自相关误差分别降低了6.1%、17.0%、28.0%和30.1%。该方法并未在所有非私有或离散基线上占优，且校正配对检验在五个种子下不显著。金丝雀暴露为1.8%，而DP Score在相同压力测试中为28.8%。这些发现支持一个透明的隐私-效用-公平性评估协议，而非临床有效性或无条件发布安全性，并推动在真实多中心记录上进行受治理的外部验证。
英文摘要
Sharing irregular longitudinal health records can accelerate model development, yet synthetic releases may leak participation, distort temporal dependence, suppress rare events, or reduce utility for underrepresented groups. We present TRUST LONGSYNTH, an auditable patient level private generator that combines bounded sufficient statistics, zCDP accounted Gaussian releases, conditional analytic score diffusion, block banded temporal covariance, separate missingness and gap models, and a protected event sampling floor with population weights. The method was evaluated on five independently generated, three cohort benchmarks containing 720 patients, fourteen irregular observation slots, six mixed variables, informative missingness, and a rare deterioration outcome. At epsilon = 12 and delta = 10 to the power of minus 5, TRUST LONGSYNTH achieved mean train synthetic test real AUPRC 0.342, Brier score 0.088, expected calibration error 0.082, correlation error 0.222, autocorrelation error 0.317, and membership attack AUROC 0.499. Relative to the private diagonal score baseline, AUPRC increased by 7.5 percent, while Brier, calibration, correlation, and autocorrelation errors decreased by 6.1 percent, 17.0 percent, 28.0 percent, and 30.1 percent, respectively. The method did not dominate every nonprivate or discrete baseline, and corrected paired tests were inconclusive with five seeds. Canary exposure was 1.8 percent, compared with 28.8 percent for DP Score in the same stress test. These findings support a transparent privacy utility fairness evaluation protocol, not clinical validity or unconditional release safety, and motivate governed external validation on real multi site records.
