Joint Entity Extraction and Assertion Detection for Clinical Text

Abstract

Negative medical findings are prevalent in clinical reports, yet discriminating them from positive findings remains a challenging task for information extraction. Most of the existing systems treat this task as a pipeline of two separate tasks, i.e., named entity recognition (NER) and rule-based negation detection. We consider this as a multi-task problem and present a novel end-to-end neural model to jointly extract entities and negations. We extend a standard hierarchical encoder-decoder NER model and first adopt a shared encoder followed by separate decoders for the two tasks. This architecture performs considerably better than the previous rule-based and machine learning-based systems. To overcome the problem of increased parameter size especially for low-resource settings, we propose the Conditional Softmax Shared Decoder architecture which achieves state-of-art results for NER and negation detection on the 2010 i2b2/VA challenge dataset and a proprietary de-identified clinical dataset.

arXiv commentAccepted at the 57th Annual Meeting of the Association for Computational Linguistics (ACL 2019)

References in corpus (8)

• Adam: A Method for Stochastic Optimization
• Practical Bayesian Optimization of Machine Learning Algorithms
• Enriching Word Vectors with Subword Information
• Multi-Task Cross-Lingual Sequence Tagging from Scratch
• NegBio: a high-performance tool for negation and uncertainty detection in radiology reports
• Improving Hospital Mortality Prediction with Medical Named Entities and Multimodal Learning
• Bidirectional LSTM-CRF for Clinical Concept Extraction
• Bidirectional LSTM-CRF for Clinical Concept Extraction

Cited by in corpus (7)

• Extracting UMLS Concepts from Medical Text Using General and Domain-Specific Deep Learning Models
• Assessment of Amazon Comprehend Medical: Medication Information Extraction
• Relation Extraction using Explicit Context Conditioning
• Neural Entity Recognition with Gazetteer based Fusion
• Cost-Quality Adaptive Active Learning for Chinese Clinical Named Entity Recognition
• To Schedule or not to Schedule: Extracting Task Specific Temporal Entities and Associated Negation Constraints
• Scope resolution of predicted negation cues: A two-step neural network-based approach