Abstract
Natural language processing (NLP) has been recently used to extract clinical information from free text in Electronic Health Record (EHR). In clinical NLP one challenge is that the meaning of clinical entities is heavily affected by assertion modifiers such as negation, uncertain, hypothetical, experiencer and so on.Incorrect assertion assignment could cause inaccurate diagnosis of patients’ condition or negatively influence following study like disease modelling. Thus, high-performance clinical NLP systems which can automatically detect negation and other assertion status of given target medical findings (e.g. disease,symptom) in clinical context are highly demanded. Here in this work, we propose a deep-learning system based on word embedding and Attention-based idirectional Long Short-Term Memory networks (AttBiLSTM) for assertion detection in clinical notes. Unlike previous state-of-art methods which requireknowledge input, our system is a knowledge poor machine learning system and can be easily extended or transferred to other domains. The evaluation of our system on public benchmarking corpora demonstrates that a knowledge poor deep-learning system can also achieve high performance for detecting negation and assertions comparing to state-of-the-art systems
Highlights
IntroductionA lot of valuable information are contained in clinical notes (e.g. patient medical history, discharge summaries, radiology reports and laboratory test results) of Electronic health records (EHRs) which can be used for various applications such as clinical decision support, disease modelling, medical risk evaluation, medication reconciliation, and quality measurements[1]
A lot of valuable information are contained in clinical notes of Electronic health records (EHRs) which can be used for various applications such as clinical decision support, disease modelling, medical risk evaluation, medication reconciliation, and quality measurements[1]
We show that a knowledge-pool deep learning system based on Att-BiLSTM networks can achieve good performance compared to state-of-art systems even with relatively small training dataset
Summary
A lot of valuable information are contained in clinical notes (e.g. patient medical history, discharge summaries, radiology reports and laboratory test results) of Electronic health records (EHRs) which can be used for various applications such as clinical decision support, disease modelling, medical risk evaluation, medication reconciliation, and quality measurements[1]. Those clinical notes which are unstructured and in free text format, are difficult and time consuming for humans to manually review or analyse. These assertion types are: absent (negation), hypothetical, possible (uncertainty), conditional
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have