A Joint Neural Baseline for Concept, Assertion, and Relation Extraction from Clinical Text

2026-03-08

AI Summary

Existing clinical information extraction methods typically treat concept recognition, assertion classification, and relation extraction as separate tasks, hindering end-to-end joint optimization. This work proposes the first unified three-stage joint modeling framework, establishing a reproducible strong baseline system that enables multi-task end-to-end training under a consistent evaluation setting. By integrating word embeddings with both general and domain-specific contextual embeddings, the model achieves substantial performance gains: F1 scores improve by 0.3, 1.4, and 3.1 points on concept recognition, assertion classification, and relation extraction, respectively, significantly outperforming conventional pipeline approaches. This advancement underscores the effectiveness of joint modeling in clinical natural language processing and paves the way for broader research and application in this domain.

Technology Category

• Natural Language Processing: Information Extraction
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• 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
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Clinical information extraction (e.g., 2010 i2b2/VA challenge) usually presents tasks of concept recognition, assertion classification, and relation extraction. Jointly modeling the multi-stage tasks in the clinical domain is an underexplored topic. The existing independent task setting (reference inputs given in each stage) makes the joint models not directly comparable to the existing pipeline work. To address these issues, we define a joint task setting and propose a novel end-to-end system to jointly optimize three-stage tasks. We empirically investigate the joint evaluation of our proposal and the pipeline baseline with various embedding techniques: word, contextual, and in-domain contextual embeddings. The proposed joint system substantially outperforms the pipeline baseline by +0.3, +1.4, +3.1 for the concept, assertion, and relation F1. This work bridges joint approaches and clinical information extraction. The proposed approach could serve as a strong joint baseline for future research. The code is publicly available.

Problem

• clinical information extraction
• concept recognition
• assertion classification
• relation extraction
• joint modeling

Innovation

• joint modeling
• clinical information extraction
• end-to-end system
• assertion classification
• relation extraction