Extracting Post-Acute Sequelae of SARS-CoV-2 Infection Symptoms from Clinical Notes via Hybrid Natural Language Processing

2025-08-17

AI Summary

Post-acute sequelae of SARS-CoV-2 infection (PASC) exhibit high symptom heterogeneity and temporal dynamics, hindering accurate clinical identification from unstructured electronic health records. Method: We developed an end-to-end hybrid NLP pipeline integrating rule-based named entity recognition with a fine-tuned BERT model for assertion classification, augmented by clinical text normalization and a curated PASC-specific terminology dictionary. Results: The system achieved an F1-score of 0.82 in single-center validation and 0.76 in ten-center external validation, with an average processing time of 2.45 seconds per note; assertion outputs showed strong correlation with ground-truth annotations (Spearman ρ > 0.83, *P* < 0.0001). Its key contribution is the first synergistic integration of structured linguistic rules and deep learning–based assertion modeling for PASC symptom extraction—enhancing both cross-center generalizability and clinical interpretability, thereby enabling robust large-scale PASC epidemiological studies.

Technology Category

• Natural Language Processing: Information Extraction
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

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

Abstract

Accurately and efficiently diagnosing Post-Acute Sequelae of COVID-19 (PASC) remains challenging due to its myriad symptoms that evolve over long- and variable-time intervals. To address this issue, we developed a hybrid natural language processing pipeline that integrates rule-based named entity recognition with BERT-based assertion detection modules for PASC-symptom extraction and assertion detection from clinical notes. We developed a comprehensive PASC lexicon with clinical specialists. From 11 health systems of the RECOVER initiative network across the U.S., we curated 160 intake progress notes for model development and evaluation, and collected 47,654 progress notes for a population-level prevalence study. We achieved an average F1 score of 0.82 in one-site internal validation and 0.76 in 10-site external validation for assertion detection. Our pipeline processed each note at $2.448pm 0.812$ seconds on average. Spearman correlation tests showed $ρ>0.83$ for positive mentions and $ρ>0.72$ for negative ones, both with $P <0.0001$. These demonstrate the effectiveness and efficiency of our models and their potential for improving PASC diagnosis.

Problem

• Extracting PASC symptoms from clinical notes accurately
• Developing hybrid NLP for symptom and assertion detection
• Improving PASC diagnosis efficiency with validated models

Innovation

• Hybrid NLP pipeline for symptom extraction
• BERT-based assertion detection modules
• Comprehensive PASC lexicon with specialists