De-identification is not always enough

no code implementations • 31 Jan 2024 • Atiquer Rahman Sarkar, Yao-Shun Chuang, Noman Mohammed, Xiaoqian Jiang

In this work, we demonstrated that (i) de-identification of real clinical notes does not protect records against a membership inference attack, (ii) proposed a novel approach to generate synthetic clinical notes using the current state-of-the-art large language models, (iii) evaluated the performance of the synthetically generated notes in a clinical domain task, and (iv) proposed a way to mount a membership inference attack where the target model is trained with synthetic data.

De-identification Inference Attack +2