ScholarDeck

AI / informaticsOphthalmology science · 2026

Automated Review of Patient Records: Privacy-Preserving Large Language Models for Identifying Incident Nonarteritic Anterior Ischemic Optic Neuropathy at Scale.

Tuyet Thao Nguyen, Kelvin Zhenghao Li, Pareena Chaitanuwong, Heather Elspeth Moss

OphthalmologyAI / informatics

Privacy-preserving large language models accurately identify nonarteritic anterior ischemic optic neuropathy.

Automated Review of Patient Records: Privacy-Preserving Large Language Models for Identifying Incident Nonarteritic Anterior Ischemic Optic Neuropathy at Scale.

Tuyet Thao Nguyen et al. · Ophthalmology science · 2026

Purpose

The purpose of this study is to evaluate automated methods for retrospective identification of acute NAION cases using large language models (LLMs) that preserve patient privacy.

Methods

Retrospective cross-sectional study.

n = 165 patients

0.85

Results

PPV

0.85 positive predictive value means 85% of identified cases were truly nonarteritic anterior ischemic optic neuropathy.

n = 165 patients

More results

7/17 prompt refinement subjects and 58/148 testing subjects had acute NAION by expert chart review corresponding to PPV of 0.39 for ≥1 ICD code.

Large language model approaches accurately identified 20 ± 12 (mean, standard deviation) acute NAION cases in the test set with PPV of 0.78 ± 0.16 and accuracy of 0.69 ± 0.06.

“

Conclusion · 1 of 2

Privacy-preserving agentic LLM approaches can achieve high PPV for acute NAION case identification using unstructured ophthalmology longitudinal electronic health records.

Conclusion · 2 of 2

These results exceed the performance of using structured ICD codes to identify cases, offering a scalable, efficient method for case identification in retrospective research while maintaining patient confidentiality and local data control.

Post

Automated Review of Patient Records: Privacy-Preserving Large Language Models for Identifying Incident Nonarteritic Anterior Ischemic Optic Neuropathy at Scale.

Privacy-preserving large language models accurately identify nonarteritic anterior ischemic optic neuropathy.

Related papers