Knowledge-Guided Retrieval-Augmented Generation for Zero-Shot Psychiatric Data: Privacy Preserving Synthetic Data Generation

2026-03-26

AI Summary

This study addresses the persistent challenges in psychiatric AI research stemming from the difficulty of accessing real patient data and associated privacy risks. To overcome these limitations, the authors propose a novel clinical knowledge–guided retrieval-augmented generation (RAG) framework that leverages the DSM-5 and ICD-10 diagnostic criteria to steer large language models in generating structurally faithful synthetic tabular data for psychiatric disorders—without requiring any real patient records. Evaluated on six anxiety disorder categories, the method achieves pairwise structural fidelity comparable to or better than established generative baselines such as CTGAN and TVAE, while substantially mitigating privacy leakage risks through its zero-shot, knowledge-driven design.

Technology Category

• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Machine Learning: Privacy
• Humans and AI: Other Foundations of Human Computation & AI

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

AI systems in healthcare research have shown potential to increase patient throughput and assist clinicians, yet progress is constrained by limited access to real patient data. To address this issue, we present a zero-shot, knowledge-guided framework for psychiatric tabular data in which large language models (LLMs) are steered via Retrieval-Augmented Generation using the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) and the International Classification of Diseases (ICD-10). We conducted experiments using different combinations of knowledge bases to generate privacy-preserving synthetic data. The resulting models were benchmarked against two state-of-the-art deep learning models for synthetic tabular data generation, namely CTGAN and TVAE, both of which rely on real data and therefore entail potential privacy risks. Evaluation was performed on six anxiety-related disorders: specific phobia, social anxiety disorder, agoraphobia, generalized anxiety disorder, separation anxiety disorder, and panic disorder. CTGAN typically achieves the best marginals and multivariate structure, while the knowledge-augmented LLM is competitive on pairwise structure and attains the lowest pairwise error in separation anxiety and social anxiety. An ablation study shows that clinical retrieval reliably improves univariate and pairwise fidelity over a no-retrieval LLM. Privacy analyses indicate that the real data-free LLM yields modest overlaps and a low average linkage risk comparable to CTGAN, whereas TVAE exhibits extensive duplication despite a low k-map score. Overall, grounding an LLM in clinical knowledge enables high-quality, privacy-preserving synthetic psychiatric data when real datasets are unavailable or cannot be shared.

Problem

• synthetic data generation
• privacy preservation
• psychiatric data
• zero-shot learning
• healthcare AI

Innovation

• Retrieval-Augmented Generation
• Zero-Shot Synthetic Data
• Privacy-Preserving AI
• Knowledge-Guided LLM
• Psychiatric Tabular Data