A long journey to short abbreviations: developing an open-source framework for clinical abbreviation recognition and disambiguation (CARD)

Wu, Yonghui; Denny, Joshua C.; Rosenbloom, S. Trent; Miller, Randolph A.; Giuse, Dario A.; Wang, Lulu; Blanquicett, Carmelo; Soysal, Ergin; Xu, Jun; Xu, Hua

doi:10.1093/jamia/ocw109

Cited by 55 publications

(47 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…al. [33]), many of which are made freely available and open source, have been intensively investigated in mining free-text medical records [10,[34][35][36]. To provide guidance in the efficient reuse of pre-trained NLP models, we have here proposed an approach that can automatically (i) identify easy cases in a new task for the reused model, on which it can achieve good performance with high confidence; (ii) classify the remainder of the cases so that the validation or retraining on them can be conducted much more efficiently, compared to adapting the model on all cases.…”

Section: Principal Resultsmentioning

confidence: 99%

Efficient Reuse of Natural Language Processing Models for Phenotype-Mention Identification in Free-text Electronic Medical Records: A Phenotype Embedding Approach

Wu¹,

Hodgson²,

Dyson³

et al. 2019

JMIR Med Inform

View full text Add to dashboard Cite

Background:Many efforts have been put into the use of automated approaches, such as natural language processing (NLP), to mine or extract data from free-text medical records to construct comprehensive patient profiles for delivering better health-care. Reusing NLP models in new settings, however, remains cumbersome -requiring validation and/or retraining on new data iteratively to achieve convergent results. Objective: The aim of this work is to minimize the effort involved in reusing NLP models on free-text medical records. Methods: We formally define and analyse the model adaptation problem in phenotype-mention identification tasks. We identify "duplicate waste" and "imbalance waste", which collectively impede efficient model reuse. We propose a phenotype embedding based approach to minimize these sources of waste without the need for labelled data from new settings. Results:We conduct experiments on data from a large mental health registry to reuse NLP models in four phenotype-mention identification tasks. The proposed approach can choose the best model for a new task, identifying up to 76% (duplicate waste), i.e. phenotype mentions without the need for validation and model retraining, and with very good performance (93-97% accuracy). It can also provide guidance for validating and retraining the selected model for novel language patterns in new tasks, saving around 80% (imbalance waste), i.e. the effort required in "blind" model-adaptation approaches. Conclusions: Adapting pre-trained NLP models for new tasks can be more efficient and effective if the language pattern landscapes of old settings and new settings can be made explicit and comparable. Our experiments show that the phenotypemention embedding approach is an effective way to model language patterns for phenotype-mention identification tasks and that its use can guide efficient NLP model reuse.

show abstract

Section: Principal Resultsmentioning

confidence: 99%

Efficient Reuse of Natural Language Processing Models for Phenotype-Mention Identification in Free-text Electronic Medical Records: A Phenotype Embedding Approach

Wu¹,

Hodgson²,

Dyson³

et al. 2019

JMIR Med Inform

View full text Add to dashboard Cite

show abstract

“…Type of abbreviation Search rule Search scope [14] Acronym LCS of initial Any subsequence in ten words before the acronym [15] Acronym LCS of initial, heuristics rules Any subsequence in the sentence [16] Acronym SVM Any sequences in the sentence exceed the number of characters in the acronym in length [17] Acronym LCS and the same initials min(|A| + 5, |A| * 2) [18] Acronym CRF Any subsequence in the sentence [19] Acronym LNCRF All the subsequence in the document [20] Abbreviation Semantic distance of context All the words in the text the performance of the normalizer is directly tied to the performance of a downstream dependency parser. In [28], Wu Y et al presented a framework called CARD(clinical abbreviation recognition and disambiguation) to handle the abbreviation in clinical data that leverages previously developed methods, including:…”

Section: Referencesmentioning

confidence: 99%

“…For this experiment, we leveraged the ShARe (Shared Annotated Resources) corpus, a subset of de-identified discharge summary, electrocardiogram, echocardiogram, and radiology reports from about 30,000 ICU (Intensive Care Unit) patients provided by the MIMIC (Multiparameter Intelligent Monitoring in Intensive Care) [44]. In this experiment, a data set with 80 clinical texts and 3024 abbreviations is selected from the corpus and three recent works (CARD [28], UTHealthCCB [29], LIMSI [30]) are used to normalize the text. The same measurements as Section 5.3 are used to evaluate the result.…”

Section: Text Normalization Improvementmentioning

confidence: 99%

Language model-based automatic prefix abbreviation expansion method for biomedical big data analysis

Zhu

et al. 2019

Future Generation Computer Systems

View full text Add to dashboard Cite

“…The biomedical abbreviations are extracted through the link-topic model algorithm (statistical model) [4]. In addition, Yonghui et al were organized open-source framework for clinical abbreviation recognition and disambiguation (including ML, clustering, and KE) [14]. De-identification is the process of removing private health information from medical discharge records.…”

Section: Applications Of Clinical Nermentioning

confidence: 99%

Evaluation of Named Entity Recognition Algorithms Using Clinical Text Data

Manimaran¹,

Velmurugan²

2018

IJET

View full text Add to dashboard Cite

Named Entity Recognition (NER) is one of the most important research areas in the field of medical. Presently, most of the clinical NER research is based on two approaches as Knowledge Engineering (KE) and Machine Learning (ML). KE is used a word lookup table approach and ML is known as supervised learning approach. The aim of this work is to evaluate a recent algorithm in KE and ML approaches using various clinical text databases. Therefore, the NOBLE Coder and Clinical Named Entity Recognition (CliNER) algorithms are selected, NOBLE Coder is depended on KE approach and CliNER is ML approach. The two algorithms will be described and compared its performance on three openly available datasets that is obtained from Medical Information Mart for Intensive Care II (MIM-IC II), Pittsburgh Medical Center, and i2b2 2010 challenge. Among these datasets, the annotated data are included which is used to detect the highest sensitivity and specificity on each algorithm. The randomly distributed patient reports were taken as input data to these algorithms. By executing these algorithms, the information is extracted and which classified into predefined concept types, for example medical problems, treatments and tests. The accuracy of both algorithms is calculated using standard measures. The taken two algorithms are analyzed based on the produced results. Finally, the best among two is suggested for better use in clinical data.

show abstract

A long journey to short abbreviations: developing an open-source framework for clinical abbreviation recognition and disambiguation (CARD)

Cited by 55 publications

References 21 publications

Efficient Reuse of Natural Language Processing Models for Phenotype-Mention Identification in Free-text Electronic Medical Records: A Phenotype Embedding Approach

Efficient Reuse of Natural Language Processing Models for Phenotype-Mention Identification in Free-text Electronic Medical Records: A Phenotype Embedding Approach

Language model-based automatic prefix abbreviation expansion method for biomedical big data analysis

Evaluation of Named Entity Recognition Algorithms Using Clinical Text Data

Contact Info

Product

Resources

About