Error Correction in Radiology Reports: A Knowledge Distillation-Based Multi-Stage Framework

2024-06-21

AI Summary

Radiology reports are prone to transcription errors, terminology inconsistencies, and template mismatches due to high clinical workload and domain-specific linguistic complexity, thereby compromising diagnostic safety. To address this, we propose a three-stage (detection–localization–correction) automated proofreading framework that emulates expert radiologist review. Our approach introduces a novel dual-knowledge injection mechanism: (1) medical knowledge graph distillation and (2) external knowledge retrieval. We further design a multi-stage decoupled proofreading paradigm tailored to real-world error patterns, integrating knowledge distillation, graph-based reasoning, retrieval-augmented generation, and sequential LLM fine-tuning. Evaluated on a real-world radiology error benchmark, our method achieves a 31.56% improvement in error detection accuracy and reduces processing latency by 37.4%. Clinical evaluation by board-certified radiologists confirms its superior clinical relevance and factual consistency over existing approaches.

Technology Category

• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Data Mining & Knowledge Management: Representing, Reasoning, and Using Provenance, Trust
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

The increasing complexity and workload of clinical radiology leads to inevitable oversights and mistakes in their use as diagnostic tools, causing delayed treatments and sometimes life-threatening harm to patients. While large language models (LLMs) have shown remarkable progress in many tasks, their utilities in detecting and correcting errors in radiology reporting are limited. This paper proposes a novel dual-knowledge infusion framework that enhances LLMs'capability for radiology report proofreading through systematic integration of medical expertise. Specifically, the knowledge infusion combines medical knowledge graph distillation (MKGD) with external knowledge retrieval (EXKR), enabling an effective automated approach in tackling mistakes in radiology reporting. By decomposing the complex proofreading task into three specialized stages of detection, localization, and correction, our method mirrors the systematic review process employed by expert radiologists, ensuring both precision and clinical interpretability. To perform a robust, clinically relevant evaluation, a comprehensive benchmark is also proposed using real-world radiology reports with real-world error patterns, including speech recognition confusions, terminology ambiguities, and template-related inconsistencies. Extensive evaluations across multiple LLM architectures demonstrate substantial improvements of our approach: up to 31.56% increase in error detection accuracy and 37.4% reduction in processing time. Human evaluation by radiologists confirms superior clinical relevance and factual consistency compared to existing approaches.

Problem

• Detecting and correcting errors in radiology reports to prevent diagnostic mistakes
• Addressing speech recognition errors and terminology ambiguities in clinical documentation
• Improving automated proofreading of radiology reports using medical knowledge integration

Innovation

• Knowledge distillation framework enhances radiology report proofreading
• Multi-stage system detects, localizes, and corrects report errors
• Combines medical knowledge graphs with external knowledge retrieval