Artificial Intelligence in Extracting Diagnostic Data from Dental Records

2024-07-23

arXiv.org

AI Summary

Periodontal disease classification updates have led to missing diagnostic information in unstructured electronic health record (EHR) text, hindering structured data extraction. To address this, we propose a hybrid diagnostic information extraction method integrating large language models (LLMs) with domain-specific fine-tuning: we first synthesize dental clinical notes using GPT-4—novel for periodontal applications—and employ them to fine-tune RoBERTa; then, we jointly perform medical named entity recognition and relation extraction to achieve end-to-end staging, grading, and subtyping of periodontitis. Our approach overcomes key challenges of low-resource settings and domain-specific terminology comprehension, while demonstrating strong cross-institutional generalizability. Evaluated on 120 real-world clinical notes from two independent centers, the method achieves overall diagnostic accuracy of 0.99 (Site 1) and 0.98 (Site 2), with perfect subtyping accuracy (1.00) at Site 2.

Technology Category

• Natural Language Processing: Information Extraction
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

This research addresses the issue of missing structured data in dental records by extracting diagnostic information from unstructured text. The updated periodontology classification system's complexity has increased incomplete or missing structured diagnoses. To tackle this, we use advanced AI and NLP methods, leveraging GPT-4 to generate synthetic notes for fine-tuning a RoBERTa model. This significantly enhances the model's ability to understand medical and dental language. We evaluated the model using 120 randomly selected clinical notes from two datasets, demonstrating its improved diagnostic extraction accuracy. The results showed high accuracy in diagnosing periodontal status, stage, and grade, with Site 1 scoring 0.99 and Site 2 scoring 0.98. In the subtype category, Site 2 achieved perfect scores, outperforming Site 1. This method enhances extraction accuracy and broadens its use across dental contexts. The study underscores AI and NLP's transformative impact on healthcare delivery and management. Integrating AI and NLP technologies enhances documentation and simplifies administrative tasks by precisely extracting complex clinical information. This approach effectively addresses challenges in dental diagnostics. Using synthetic training data from LLMs optimizes the training process, improving accuracy and efficiency in identifying periodontal diagnoses from clinical notes. This innovative method holds promise for broader healthcare applications, potentially improving patient care quality.

Problem

• Extracting dental diagnoses from unstructured clinical notes
• Improving accuracy in periodontal status classification
• Enhancing cross-institutional dental data interoperability

Innovation

• Generative AI creates synthetic dental notes
• RoBERTa model fine-tuned with GPT-4 data
• High accuracy in periodontal diagnosis extraction