arXiv Computation and Language September 1, 2026 By Yuping Wu, Viktor Schlegel, Warren Del-Pinto, Srinivasan Nandakumar, Iqra Zahid, Yidan Sun, Hai Li, Usama Farghaly Omar, Amirah Jasmine, Arun-Kumar Kaliya-Perumal, Chun Shen Tham, Gabriel Connors, Anil A Bharath, Goran Nenadic

Privacy-Preserving Generation of Clinical Narratives from Medical Terminologies

Read the original on arXiv Computation and Language →

The paper introduces Term2Note, a method for generating full-length clinical notes under differential privacy constraints. It separates content and form, conditioning note sections on medical terms and applying distinct DP protections to terms and notes, followed by a DP quality maximizer. Experiments show that the synthetic notes closely match real clinical notes in statistical properties, and models trained on them perform comparably to those trained on real data, outperforming existing DP text generation baselines.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computation and Language.

safety

arXiv Computation and Language

Aug 25

Clinically Grounded Privacy Evaluation of Medical LMs

The paper introduces a clinically grounded privacy evaluation framework for medical language models, assessing leakage across a spectrum of adversarial access levels—from publicly inferable demographics to leaked note fragments. Using this framework on an LM pretrained on 378,000 clinical notes, the authors find that routine encounter metadata leads to high verbatim memorization and significant recovery of sensitive diagnoses (e.g., AUROC 0.91 for abortion, 0.82 for HIV). They also note that exact-match memorization can overstate disclosure, with 36% of memorized tokens being templated documentation, underscoring the risks of training on longitudinal clinical data and offering a reusable evaluation tool.

By Sasha Ronaghi, Sana Tonekaboni, Lena Stempfle, Vivian Utti, Jordan Li Cahoon, Nathaniel Hendrix, Ayin Vala, Marzyeh Ghassemi, Emily Alsentzer

llms safety

arXiv AI

Jun 26

A Pipeline for Generating Longitudinal Synthetic Clinical Notes Using Large Language Models

arXiv:2606. 26879v1 Announce Type: new Abstract: Synthetic data is increasingly used to enable the development and evaluation of AI systems in domains where access to real-world data is restricted.

By William Poulett

llms safety

Hugging Face Trending Papers

Jun 25

A Pipeline for Generating Longitudinal Synthetic Clinical Notes Using Large Language Models

Synthetic data is increasingly used to enable the development and evaluation of AI systems in domains where access to real-world data is restricted. In healthcare, clinical documentation presents particular challenges due to its sensitivity.

llms safety

arXiv AI

Aug 25

Future Querying: Can LLMs Serve as Implicit Medical World Models?

arXiv:2608.23248v1 Announce Type: cross Abstract: Traditional clinical prediction models rely on task-specific pipelines and curated, structured data, which scale poorly and underutilize unstructured...

By Siri Willems, James Butterworth, Lore Goetschalckx, Peter Vrancx, Philippe Modard, Elke Giets, Ludovic Denoyer

llms fine-tuning safety

arXiv AI

Jun 11

Self-Prompting Small Language Models for Privacy-Sensitive Clinical Information Extraction

arXiv:2605. 04221v2 Announce Type: replace-cross Abstract: Clinical named entity recognition from dental progress notes is challenging because documentation is highly unstructured, domain-specific, and often privacy-sensitive.

By Yao-Shun Chuang, Tushti Mody, Uday Pratap Singh, Shirindokht Shiraz, Chun-Teh Lee, Ryan Brandon, Muhammad F Walji, Xiaoqian Jiang, Bunmi Tokede

llms nlp reinforcement-learning fine-tuning benchmarks safety

arXiv AI

Jun 16

MedSynth: Realistic, Synthetic Medical Dialogue-Note Pairs

arXiv:2508. 01401v2 Announce Type: replace-cross Abstract: Physicians spend significant time documenting clinical encounters, a burden that contributes to professional burnout.

By Ahmad Rezaie Mianroodi, Amirali Rezaie, Niko Grisel Todorov, Nadine A. Friedrich, Maria P Mogollon, Alexander Hernandez-Tirado, Guillermo Lopez Garcia, Cyril Rakovski, Frank Rudzicz

safety

One email a morning, machine-written

Related stories

Clinically Grounded Privacy Evaluation of Medical LMs

A Pipeline for Generating Longitudinal Synthetic Clinical Notes Using Large Language Models

A Pipeline for Generating Longitudinal Synthetic Clinical Notes Using Large Language Models

Future Querying: Can LLMs Serve as Implicit Medical World Models?

Self-Prompting Small Language Models for Privacy-Sensitive Clinical Information Extraction

MedSynth: Realistic, Synthetic Medical Dialogue-Note Pairs