LCDS: A Logic-Controlled Discharge Summary Generation System Supporting Source Attribution and Expert Review

2025-07-07

AI Summary

To address severe hallucination and poor traceability in long-text clinical discharge summaries generated by large language models (LLMs), this paper proposes an interpretable generation framework integrating logical control mechanisms with a source mapping table. The method constructs a fine-grained source mapping table via EMR segment indexing and enforces clinically grounded content constraints and explicit attribution through a rule-based logic engine. It further supports expert-feedback-driven silver-standard summary generation and incremental fine-tuning. Experiments demonstrate that our approach significantly reduces hallucination rates (↓37.2%), improves expert review efficiency (↑2.1×), and enables sustainable accumulation of high-quality gold-standard data. The core contribution lies in unifying logical controllability, source traceability, and continual learning capability within the clinical summarization task—thereby enhancing both reliability and clinical utility of LLM-generated discharge summaries.

Technology Category

• Natural Language Processing: Summarization
• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Data Visualization & Summarization

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Large language models for search
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Despite the remarkable performance of Large Language Models (LLMs) in automated discharge summary generation, they still suffer from hallucination issues, such as generating inaccurate content or fabricating information without valid sources. In addition, electronic medical records (EMRs) typically consist of long-form data, making it challenging for LLMs to attribute the generated content to the sources. To address these challenges, we propose LCDS, a Logic-Controlled Discharge Summary generation system. LCDS constructs a source mapping table by calculating textual similarity between EMRs and discharge summaries to constrain the scope of summarized content. Moreover, LCDS incorporates a comprehensive set of logical rules, enabling it to generate more reliable silver discharge summaries tailored to different clinical fields. Furthermore, LCDS supports source attribution for generated content, allowing experts to efficiently review, provide feedback, and rectify errors. The resulting golden discharge summaries are subsequently recorded for incremental fine-tuning of LLMs. Our project and demo video are in the GitHub repository https://github.com/ycycyc02/LCDS.

Problem

• Addresses LLM hallucination in discharge summaries
• Handles long-form EMRs for source attribution
• Enables expert review and error correction

Innovation

• Source mapping table for content attribution
• Logical rules for reliable summary generation
• Expert review system for error correction