Learning to Infer Entities, Properties and their Relations from Clinical Conversations

Du, Nan; Wang, Mingqiu; Tran, Linh; Lee, Gang; Shafran, Izhak

doi:10.18653/v1/d19-1503

Cited by 25 publications

(22 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…After screening the titles and abstracts of these articles, we assessed 144 full-text articles for eligibility. We included 20 articles [19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38] for our analysis (Fig. 1 and Supplementary Table 2).…”

Section: Study Selectionmentioning

confidence: 99%

The digital scribe in clinical practice: a scoping review and research agenda

et al. 2021

View full text Add to dashboard Cite

The number of clinician burnouts is increasing and has been linked to a high administrative burden. Automatic speech recognition (ASR) and natural language processing (NLP) techniques may address this issue by creating the possibility of automating clinical documentation with a “digital scribe”. We reviewed the current status of the digital scribe in development towards clinical practice and present a scope for future research. We performed a literature search of four scientific databases (Medline, Web of Science, ACL, and Arxiv) and requested several companies that offer digital scribes to provide performance data. We included articles that described the use of models on clinical conversational data, either automatically or manually transcribed, to automate clinical documentation. Of 20 included articles, three described ASR models for clinical conversations. The other 17 articles presented models for entity extraction, classification, or summarization of clinical conversations. Two studies examined the system’s clinical validity and usability, while the other 18 studies only assessed their model’s technical validity on the specific NLP task. One company provided performance data. The most promising models use context-sensitive word embeddings in combination with attention-based neural networks. However, the studies on digital scribes only focus on technical validity, while companies offering digital scribes do not publish information on any of the research phases. Future research should focus on more extensive reporting, iteratively studying technical validity and clinical validity and usability, and investigating the clinical utility of digital scribes.

show abstract

Section: Study Selectionmentioning

confidence: 99%

The digital scribe in clinical practice: a scoping review and research agenda

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Moreover, unlike a span extraction model, the generative model might produce hallucinated facts. Du et al (2019) obtain MR attributes as spans in text; however, they use a fully supervised approach which requires a large dataset with spanlevel labels.…”

Section: Related Workmentioning

confidence: 99%

“…In this work, we focus on extracting Medication Regimen (MR) information (Du et al, 2019;Selvaraj and Konam, 2019) from the doctor-patient conversations. Specifically, we extract three attributes, i.e., frequency, route and change, corresponding to medications discussed in the conversation (Figure 1).…”

Section: Introductionmentioning

confidence: 99%

Weakly Supervised Medication Regimen Extraction from Medical Conversations

Patel

Konam

Prabhakar³

2020

Proceedings of the 3rd Clinical Natural Language Processing Workshop

View full text Add to dashboard Cite

Automated Medication Regimen (MR) extraction from medical conversations can not only improve recall and help patients follow through with their care plan, but also reduce the documentation burden for doctors. In this paper, we focus on extracting spans for frequency, route and change, corresponding to medications discussed in the conversation. We first describe a unique dataset of annotated doctor-patient conversations and then present a weakly supervised model architecture that can perform span extraction using noisy classification data. The model utilizes an attention bottleneck inside a classification model to perform the extraction. We experiment with several variants of attention scoring and projection functions and propose a novel transformer-based attention scoring function (TAScore). The proposed combination of TAScore and Fusedmax projection achieves a 10 point increase in Longest Common Substring F1 compared to the baseline of additive scoring plus softmax projection.

show abstract

“…The authors also employed a transformer network [ 25 ] for building an intersentence model. For clinical conversations, Du et al [ 26 ] proposed a relation span attribute tagging (R-SAT) model that utilizes bi-LSTM and has been shown to outperform the baseline by a large margin for two relation classification tasks.…”

Section: Introductionmentioning

confidence: 99%

Relation Classification for Bleeding Events From Electronic Health Records Using Deep Learning Systems: An Empirical Study

et al. 2021

View full text Add to dashboard Cite

Background Accurate detection of bleeding events from electronic health records (EHRs) is crucial for identifying and characterizing different common and serious medical problems. To extract such information from EHRs, it is essential to identify the relations between bleeding events and related clinical entities (eg, bleeding anatomic sites and lab tests). With the advent of natural language processing (NLP) and deep learning (DL)-based techniques, many studies have focused on their applicability for various clinical applications. However, no prior work has utilized DL to extract relations between bleeding events and relevant entities. Objective In this study, we aimed to evaluate multiple DL systems on a novel EHR data set for bleeding event–related relation classification. Methods We first expert annotated a new data set of 1046 deidentified EHR notes for bleeding events and their attributes. On this data set, we evaluated three state-of-the-art DL architectures for the bleeding event relation classification task, namely, convolutional neural network (CNN), attention-guided graph convolutional network (AGGCN), and Bidirectional Encoder Representations from Transformers (BERT). We used three BERT-based models, namely, BERT pretrained on biomedical data (BioBERT), BioBERT pretrained on clinical text (Bio+Clinical BERT), and BioBERT pretrained on EHR notes (EhrBERT). Results Our experiments showed that the BERT-based models significantly outperformed the CNN and AGGCN models. Specifically, BioBERT achieved a macro F1 score of 0.842, outperforming both the AGGCN (macro F1 score, 0.828) and CNN models (macro F1 score, 0.763) by 1.4% (P<.001) and 7.9% (P<.001), respectively. Conclusions In this comprehensive study, we explored and compared different DL systems to classify relations between bleeding events and other medical concepts. On our corpus, BERT-based models outperformed other DL models for identifying the relations of bleeding-related entities. In addition to pretrained contextualized word representation, BERT-based models benefited from the use of target entity representation over traditional sequence representation

show abstract

Learning to Infer Entities, Properties and their Relations from Clinical Conversations

Cited by 25 publications

References 37 publications

The digital scribe in clinical practice: a scoping review and research agenda

The digital scribe in clinical practice: a scoping review and research agenda

Weakly Supervised Medication Regimen Extraction from Medical Conversations

Relation Classification for Bleeding Events From Electronic Health Records Using Deep Learning Systems: An Empirical Study

Contact Info

Product

Resources

About