TraceCoder: Towards Traceable ICD Coding via Multi-Source Knowledge Integration

2025-10-16

AI Summary

Existing ICD automatic coding methods face three major challenges: substantial semantic gaps between clinical text and codes, weak performance on long-tail and rare codes, and poor model interpretability. To address these, this paper proposes a knowledge-enhanced dynamic ensemble framework—the first to jointly integrate UMLS semantic mappings, Wikipedia-based contextual expansion, and large language model (LLM) representations. A hybrid attention mechanism is designed to enable fine-grained, interactive modeling among clinical notes, ICD labels, and heterogeneous external knowledge sources, thereby significantly improving prediction traceability and interpretability. Evaluated on MIMIC-III and MIMIC-IV, our method achieves state-of-the-art performance, with F1-score gains of 3.2–5.7% overall and a remarkable +12.4% improvement for rare codes. Ablation studies confirm the efficacy of each component. The framework delivers high accuracy, clinically meaningful explanations, and practical deployability.

Technology Category

• Machine Learning: Ensemble Methods
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Automated International Classification of Diseases (ICD) coding assigns standardized diagnosis and procedure codes to clinical records, playing a critical role in healthcare systems. However, existing methods face challenges such as semantic gaps between clinical text and ICD codes, poor performance on rare and long-tail codes, and limited interpretability. To address these issues, we propose TraceCoder, a novel framework integrating multi-source external knowledge to enhance traceability and explainability in ICD coding. TraceCoder dynamically incorporates diverse knowledge sources, including UMLS, Wikipedia, and large language models (LLMs), to enrich code representations, bridge semantic gaps, and handle rare and ambiguous codes. It also introduces a hybrid attention mechanism to model interactions among labels, clinical context, and knowledge, improving long-tail code recognition and making predictions interpretable by grounding them in external evidence. Experiments on MIMIC-III-ICD9, MIMIC-IV-ICD9, and MIMIC-IV-ICD10 datasets demonstrate that TraceCoder achieves state-of-the-art performance, with ablation studies validating the effectiveness of its components. TraceCoder offers a scalable and robust solution for automated ICD coding, aligning with clinical needs for accuracy, interpretability, and reliability.

Problem

• Bridging semantic gaps between clinical text and ICD codes
• Improving performance on rare and long-tail ICD codes
• Enhancing interpretability of automated ICD coding predictions

Innovation

• Integrates UMLS, Wikipedia, and LLMs for knowledge
• Uses hybrid attention to model label-context interactions
• Enhances rare code recognition and interpretable predictions