Unified Medical Language System Concepts Research Articles

Streamlining social media information extraction for public health research with deep learning.

Social media-based public health research is crucial for epidemic surveillance, but most studies identify relevant corpora with keyword-matching. This study develops a system to streamline the process of curating colloquial medical dictionaries. We demonstrate the pipeline by curating a UMLS-colloquial symptom dictionary from COVID-19-related tweets as proof of concept. COVID-19-related tweets from February 1, 2020, to April 30, 2022 were used. The pipeline includes three modules: a named entity recognition module to detect symptoms in tweets; an entity normalization module to aggregate detected entities; and a mapping module that iteratively maps entities to Unified Medical Language System concepts. A random 500 entity sample were drawn from the final dictionary for accuracy validation. Additionally, we conducted a symptom frequency distribution analysis to compare our dictionary to a pre-defined lexicon from previous research. We identified 498,480 unique symptom entity expressions from the tweets. Pre-processing reduces the number to 18,226. The final dictionary contains 38,175 unique expressions of symptoms that can be mapped to 966 UMLS concepts (accuracy = 95%). Symptom distribution analysis found that our dictionary detects more symptoms and is effective at identifying psychiatric disorders like anxiety and depression, often missed by pre-defined lexicons. This study advances public health research by implementing a novel, systematic pipeline for curating symptom lexicons from social media data. The final lexicon's high accuracy, validated by medical professionals, underscores the potential of this methodology to reliably interpret and categorize vast amounts of unstructured social media data into actionable medical insights across diverse linguistic and regional landscapes.

Representation of child and youth participation within the Unified Medical Language System (UMLS)

Purpose To examine (1) how much participation is represented in the benchmark Unified Medical Language System (UMLS) resource, and (2) to what extent that representation reflects the definition of child and youth participation and/or its related constructs per the family of Participation-Related Constructs framework. Materials and methods We searched and analysed UMLS concepts related to the term “participation.” Identified UMLS concepts were rated according to their representation of participation (i.e., attendance, involvement, both) as well as participation-related constructs using deductive content analysis. Results 363 UMLS concepts were identified. Of those, 68 had at least one English definition, resulting in 81 definitions that were further analysed. Results revealed 2 definitions (2/81; 3%; 2/68 UMLS concepts) representing participation “attendance” and 18 definitions (18/81; 22%; 14/68 UMLS concepts) representing participation “involvement.” No UMLS concept definition represented both attendance and involvement (i.e., participation). Most of the definitions (11/20; 55%; 9/16 UMLS concepts) representing attendance or involvement also represent a participation-related construct. Conclusion(s) The representation of participation within the UMLS is limited and poorly aligned with the contemporary definition of child and youth participation. Expanding ontological resources to represent child and youth participation is needed to enable better data analytics that reflect contemporary paediatric rehabilitation practice.

Complementary and Integrative Health Information in the literature: its lexicon and named entity recognition.

To construct an exhaustive Complementary and Integrative Health (CIH) Lexicon (CIHLex) to help better represent the often underrepresented physical and psychological CIH approaches in standard terminologies, and to also apply state-of-the-art natural language processing (NLP) techniques to help recognize them in the biomedical literature. We constructed the CIHLex by integrating various resources, compiling and integrating data from biomedical literature and relevant sources of knowledge. The Lexicon encompasses 724 unique concepts with 885 corresponding unique terms. We matched these concepts to the Unified Medical Language System (UMLS), and we developed and utilized BERT models comparing their efficiency in CIH named entity recognition to well-established models including MetaMap and CLAMP, as well as the large language model GPT3.5-turbo. Of the 724 unique concepts in CIHLex, 27.2% could be matched to at least one term in the UMLS. About 74.9% of the mapped UMLS Concept Unique Identifiers were categorized as "Therapeutic or Preventive Procedure." Among the models applied to CIH named entity recognition, BLUEBERT delivered the highest macro-average F1-score of 0.91, surpassing other models. Our CIHLex significantly augments representation of CIH approaches in biomedical literature. Demonstrating the utility of advanced NLP models, BERT notably excelled in CIH entity recognition. These results highlight promising strategies for enhancing standardization and recognition of CIH terminology in biomedical contexts.

Abstract 17253: Concepts Associated With 30-day Rehospitalization From Unstructured Data Among Patients With Heart Failure

Introduction: Heart failure (HF) is a diagnosis for which the Centers for Medicare & Medicaid Services mandated public reporting on 30-day risk-standardized rehospitalization rates. Examining risk factors predicting rehospitalization has been widely studied, and most of studies utilized structured data from electronic health records (EHRs) in HF. However, those studies did not use patient-level information that could be captured from unstructured data. Hypothesis: Unified Medical Language System (UMLS) concepts that may be used as factors that could be associated with the likelihood of 30-day rehospitalization among patients with HF by using natural language processing technique. Methods: We obtained notes from patients diagnosed with HF from EHRs from a tertiary hospital between June 1, 2015 and December 31, 2019. We analyzed 14 note types including physician notes, nursing notes and discharge summaries. We used the publicly available natural language processing software MetaMap to extract UMLS concepts that may be used as factors associated with the liekhood of 30-day rehospitalization and the Pandas package in Python for processing and analysis. Results: We identified 350,272 notes among 4,773 patients discharged with HF. The median age of the cohort was 76 years, 54.3% were male, and 89% white, and the 30 day-rehospitalization rate was 12.8%.. A total of 350,272 notes were for 4,164 patients who were not rehospitalized, and 57,212 notes for 609 patients who were rehospitalized. The UMLS concepts identified at a higher rate in notes of rehospitalized patients compared to non-rehospitalized patients included “personal appearance” (7.0% vs 5.5%), “power (psychology)” (3.5% vs 2.5%), and “intrinsic drive” (2.1% vs 1.1%). The UMLS Concepts that were mentioned less in rehospitalized patient notes include “family” (26.3% vs 30.8%), “home environment” (59.9% vs 64%), and “independence” (20.6% vs 24.6%). Conclusions: We identified six UMLS concepts associated with 30-day rehospitalization in patients with HF. Further study is warranted to determine whether incorporation of concepts from unstructured data could improve prediction of 30-day rehospitalization and provide opportunities for patient-level intervention.

Cerebrovascular disease case identification in inpatient electronic medical record data using natural language processing

Backgrounding cerebrovascular disease (CeVD) from inpatient electronic medical records (EMRs) through natural language processing (NLP) is pivotal for automated disease surveillance and improving patient outcomes. Existing methods rely on coders’ abstraction, which has time delays and under-coding issues. This study sought to develop an NLP-based method to detect CeVD using EMR clinical notes.MethodsCeVD status was confirmed through a chart review on randomly selected hospitalized patients who were 18 years or older and discharged from 3 hospitals in Calgary, Alberta, Canada, between January 1 and June 30, 2015. These patients’ chart data were linked to administrative discharge abstract database (DAD) and Sunrise™ Clinical Manager (SCM) EMR database records by Personal Health Number (a unique lifetime identifier) and admission date. We trained multiple natural language processing (NLP) predictive models by combining two clinical concept extraction methods and two supervised machine learning (ML) methods: random forest and XGBoost. Using chart review as the reference standard, we compared the model performances with those of the commonly applied International Classification of Diseases (ICD-10-CA) codes, on the metrics of sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).ResultOf the study sample (n = 3036), the prevalence of CeVD was 11.8% (n = 360); the median patient age was 63; and females accounted for 50.3% (n = 1528) based on chart data. Among 49 extracted clinical documents from the EMR, four document types were identified as the most influential text sources for identifying CeVD disease (“nursing transfer report,” “discharge summary,” “nursing notes,” and “inpatient consultation.”). The best performing NLP model was XGBoost, combining the Unified Medical Language System concepts extracted by cTAKES (e.g., top-ranked concepts, “Cerebrovascular accident” and “Transient ischemic attack”), and the term frequency-inverse document frequency vectorizer. Compared with ICD codes, the model achieved higher validity overall, such as sensitivity (25.0% vs 70.0%), specificity (99.3% vs 99.1%), PPV (82.6 vs. 87.8%), and NPV (90.8% vs 97.1%).ConclusionThe NLP algorithm developed in this study performed better than the ICD code algorithm in detecting CeVD. The NLP models could result in an automated EMR tool for identifying CeVD cases and be applied for future studies such as surveillance, and longitudinal studies.

The suitability of UMLS and SNOMED-CT for encoding outcome concepts.

Outcomes are important clinical study information. Despite progress in automated extraction of PICO (Population, Intervention, Comparison, and Outcome) entities from PubMed, rarely are these entities encoded by standard terminology to achieve semantic interoperability. This study aims to evaluate the suitability of the Unified Medical Language System (UMLS) and SNOMED-CT in encoding outcome concepts in randomized controlled trial (RCT) abstracts. We iteratively developed and validated an outcome annotation guideline and manually annotated clinically significant outcome entities in the Results and Conclusions sections of 500 randomly selected RCT abstracts on PubMed. The extracted outcomes were fully, partially, or not mapped to the UMLS via MetaMap based on established heuristics. Manual UMLS browser search was performed for select unmapped outcome entities to further differentiate between UMLS and MetaMap errors. Only 44% of 2617 outcome concepts were fully covered in the UMLS, among which 67% were complex concepts that required the combination of 2 or more UMLS concepts to represent them. SNOMED-CT was present as a source in 61% of the fully mapped outcomes. Domains such as Metabolism and Nutrition, and Infections and Infectious Diseases need expanded outcome concept coverage in the UMLS and MetaMap. Future work is warranted to similarly assess the terminology coverage for P, I, C entities. Computational representation of clinical outcomes is important for clinical evidence extraction and appraisal and yet faces challenges from the inherent complexity and lack of coverage of these concepts in UMLS and SNOMED-CT, as demonstrated in this study.

LeafAI: query generator for clinical cohort discovery rivaling a human programmer.

Identifying study-eligible patients within clinical databases is a critical step in clinical research. However, accurate query design typically requires extensive technical and biomedical expertise. We sought to create a system capable of generating data model-agnostic queries while also providing novel logical reasoning capabilities for complex clinical trial eligibility criteria. The task of query creation from eligibility criteria requires solving several text-processing problems, including named entity recognition and relation extraction, sequence-to-sequence transformation, normalization, and reasoning. We incorporated hybrid deep learning and rule-based modules for these, as well as a knowledge base of the Unified Medical Language System (UMLS) and linked ontologies. To enable data-model agnostic query creation, we introduce a novel method for tagging database schema elements using UMLS concepts. To evaluate our system, called LeafAI, we compared the capability of LeafAI to a human database programmer to identify patients who had been enrolled in 8 clinical trials conducted at our institution. We measured performance by the number of actual enrolled patients matched by generated queries. LeafAI matched a mean 43% of enrolled patients with 27 225 eligible across 8 clinical trials, compared to 27% matched and 14 587 eligible in queries by a human database programmer. The human programmer spent 26 total hours crafting queries compared to several minutes by LeafAI. Our work contributes a state-of-the-art data model-agnostic query generation system capable of conditional reasoning using a knowledge base. We demonstrate that LeafAI can rival an experienced human programmer in finding patients eligible for clinical trials.

Ontology-driven and weakly supervised rare disease identification from clinical notes

BackgroundComputational text phenotyping is the practice of identifying patients with certain disorders and traits from clinical notes. Rare diseases are challenging to be identified due to few cases available for machine learning and the need for data annotation from domain experts.MethodsWe propose a method using ontologies and weak supervision, with recent pre-trained contextual representations from Bi-directional Transformers (e.g. BERT). The ontology-driven framework includes two steps: (i) Text-to-UMLS, extracting phenotypes by contextually linking mentions to concepts in Unified Medical Language System (UMLS), with a Named Entity Recognition and Linking (NER+L) tool, SemEHR, and weak supervision with customised rules and contextual mention representation; (ii) UMLS-to-ORDO, matching UMLS concepts to rare diseases in Orphanet Rare Disease Ontology (ORDO). The weakly supervised approach is proposed to learn a phenotype confirmation model to improve Text-to-UMLS linking, without annotated data from domain experts. We evaluated the approach on three clinical datasets, MIMIC-III discharge summaries, MIMIC-III radiology reports, and NHS Tayside brain imaging reports from two institutions in the US and the UK, with annotations.ResultsThe improvements in the precision were pronounced (by over 30% to 50% absolute score for Text-to-UMLS linking), with almost no loss of recall compared to the existing NER+L tool, SemEHR. Results on radiology reports from MIMIC-III and NHS Tayside were consistent with the discharge summaries. The overall pipeline processing clinical notes can extract rare disease cases, mostly uncaptured in structured data (manually assigned ICD codes).ConclusionThe study provides empirical evidence for the task by applying a weakly supervised NLP pipeline on clinical notes. The proposed weak supervised deep learning approach requires no human annotation except for validation and testing, by leveraging ontologies, NER+L tools, and contextual representations. The study also demonstrates that Natural Language Processing (NLP) can complement traditional ICD-based approaches to better estimate rare diseases in clinical notes. We discuss the usefulness and limitations of the weak supervision approach and propose directions for future studies.

АВТОМАТИЗИРОВАННАЯ СИСТЕМА ИЗВЛЕЧЕНИЯ КЛИНИЧЕСКИ РЕЛЕВАНТНЫХ ТЕРМИНОВ UMLS ИЗ ТЕКСТОВ АНГЛОЯЗЫЧНЫХ СТАТЕЙ НА ПРИМЕРЕ АКСИАЛЬНОГО СПОНДИЛОАРТРИТА

Significance. Early detection of axial spondyloarthritis (axSpA) is a complex clinical task. Quality improvement of axSpA diagnostics in primary care settings is possible with the help of decision-making information systems based on the ontological approach application. The key stage of the decision-making system development consists of the elaboration of a set of clinical terms. This set should fully describe clinical area or sub-area under study. One of the essential requirements is compliance of the clinical terms used with the existing clinical nomenclatures. Currently, the largest set of clinical terms is the Unified Medical Language System (UMLS) metathesaurus. The majority of UMLS terms is presented in English only. The development of tools for the analysis of unstructured texts and recognition of clinically relevant UMLS entities make it possible to elaborate a set of terms describing axSpA diagnostic aspects. This will also help to compile a list of UMLS terminology nomenclatures for their priority adaptation and expert translation into Russian. The purpose of this study is to develop an automated system for recognizing clinically relevant UMLS terms in texts of the English-language articles. Material and methods. The research material has included English terms (11.2 million) aggregated from 76 nome6nclatures of the current UMLS (2022AB) version. In addition, the study has used texts of PubMed clinical abstracts in English. Queries to the UMLS graph model, semantic algorithms for unstructured texts and machine-learning methods have been applied for data collection and analysis. Results. The study has elaborated a set of high-accuracy regular expressions (F1-score=98%) for metadata elimination from the text corpus. Then the authors have identified patterns for searching clinically relevant terms in the aggregated set of UMLS concepts. Using a logistic regression algorithm, the authors have trained a binary classification model. Input data for the created classificatory are information about an UMLS term. Output data are a label indicating the presence or absence of clinical relevance. Conclusion. The binary classification model has been validated individually and double-tested on different data samples. Values of the accuracy, sensitivity and specificity of metrics equal to 91%, 90% and 91%, respectively, for the validation sample (a number of axSpA terms). In addition, this model has been tested on the sets of terms aggregated for any two diseases. Values of the accuracy metric equal to 91% and 90%, respectively. With the help of the developed machine-learning model, the study has estimated that UMLS contains 1.5 million unique terms applicable to describing a clinical picture. In addition, lists of priority UMLS data sources and thematic groups have been compiled. These clinically relevant UMLS terms should be adapted and translated into Russian as soon as possible.

Mapping Chinese Medical Entities to the Unified Medical Language System.

Chinese medical entities have not been organized comprehensively due to the lack of well-developed terminology systems, which poses a challenge to processing Chinese medical texts for fine-grained medical knowledge representation. To unify Chinese medical terminologies, mapping Chinese medical entities to their English counterparts in the Unified Medical Language System (UMLS) is an efficient solution. However, their mappings have not been investigated sufficiently in former research. In this study, we explore strategies for mapping Chinese medical entities to the UMLS and systematically evaluate the mapping performance. First, Chinese medical entities are translated to English using multiple web-based translation engines. Then, 3 mapping strategies are investigated: (a) string-based, (b) semantic-based, and (c) string and semantic similarity combined. In addition, cross-lingual pretrained language models are applied to map Chinese medical entities to UMLS concepts without translation. All of these strategies are evaluated on the ICD10-CN, Chinese Human Phenotype Ontology (CHPO), and RealWorld datasets. The linear combination method based on the SapBERT and term frequency-inverse document frequency bag-of-words models perform the best on all evaluation datasets, with 91.85%, 82.44%, and 78.43% of the top 5 accuracies on the ICD10-CN, CHPO, and RealWorld datasets, respectively. In our study, we explore strategies for mapping Chinese medical entities to the UMLS and identify a satisfactory linear combination method. Our investigation will facilitate Chinese medical entity normalization and inspire research that focuses on Chinese medical ontology development.

A practical approach to identifying autistic adults within the electronic health record.

The electronic health record (EHR) provides valuable data for understanding physical and mental health conditions in autism. We developed an approach to identify charts of autistic young adults, retrieved from our institution's de-identified EHR database. Clinical notes within two cohorts were identified. Cohort 1 charts had at least one International Classification of Diseases (ICD-CM) autism code. Cohort 2 charts had only autism key terms without ICD-CM codes, and at least four notes per chart. A natural language processing tool parsed medical charts to identify key terms associated with autism diagnoses and mapped them to Unified Medical Language System Concept Unique Identifiers (CUIs). Average scores were calculated for each set of charts based on captured CUIs. Chart review determined whether patients met criteria for autism using a classification rubric. In Cohort 1, of 418 patients, 361 were confirmed to have autism by chart review. Sensitivity was 0.99 and specificity was 0.68 with positive predictive value (PPV) of 0.97. Specificity improved to 0.81 (sensitivity was 0.95; PPV was 0.98) when the number of notes was limited to four or more per chart. In Cohort 2, 48 of 136 patients were confirmed to have autism by chart review. Sensitivity was 0.95, specificity was 0.73, and PPV was 0.70. Our approach, which included using key terms, identified autism charts with high sensitivity, even in the absence of ICD-CM codes. Relying on ICD-CM codes alone may result in inclusion of false positive cases and exclusion of true cases with autism.

Analysis of the Representation of Frequent Clinical Attributes in the Unified Medical Language System.

Mapping clinical attributes from hospital information systems to standardized terminologies may allow their scientific reuse for multicenter studies. The Unified Medical Language System (UMLS) defines synonyms in different terminologies, which could be valuable for achieving semantic interoperability between different sites. Here we aim to explore the potential relevance of UMLS concepts and associated semantic relations for widely used clinical terminologies in a German university hospital. To semi-automatically examine a sample of the 200 most frequent codes from Erlangen University Hospital for three relevant terminologies, we implemented a script that queries their UMLS representation and associated mappings via a programming interface. We found that 94% of frequent diagnostic codes were available in UMLS, and that most of these codes could be mapped to other terminologies such as SNOMED CT. We observed that all examined laboratory codes were represented in UMLS, and that various translations to other languages were available for these concepts. The classification that is most widely used in German hospital for documenting clinical procedures was not originally represented in UMLS, but external mappings to SNOMED CT allowed identifying UMLS entries for 90.5% of frequent codes. Future research could extend this investigation to other code sets and terminologies, or study the potential utility of available mappings for specific applications.

A configurable software platform for creating, reviewing and adjudicating annotation of unstructured text.

ObjectivesTo develop a flexible platform for creating, reviewing and adjudicating annotation of unstructured text. Natural Language Processing models and statistical classifiers use the results for analysis of large databases of text, such as electronic health records, that are curated by the National Centre for Healthy Ageing (NCHA) Data Platform.
 ApproachAutomated approaches are essential for large scale extraction of structured data from unstructured documents. We applied the CogStack suite to annotate clinical text from hospital inpatient records based on the Unified Medical Language System (UMLS) for classifying dementia status. We trained a logistic regression classifier to determine dementia/non-dementia status within two cohorts based on frequency of occurrence of a set of terms provided by experts - one with confirmed dementia based on clinical assessment and the other confirmed non-dementia based on telephone cognitive interview. We used our annotation platform to review the accuracy of concepts assigned by CogStack.
 ResultsThere were 368 people with clinically confirmed dementia and 218 screen-negative for dementia. Of these, 259 with dementia and 195 without dementia had documents in the inpatient electronic health record system, 84045 inpatient documents 16950 for the dementia and non-dementia cohort respectively. A set of key words pertaining to dementia was generated by a specialist neurologist and a health information manager, and matched to UMLS concepts. The NCHA data platform holds a copy of the inpatient text records (>13million documents) that has been annotated using CogStack. Annotated documents corresponding to the study cohort were extracted.
 We tested true positive rates of annotation against 50 concepts judged by a neurologist and health information manager to be relevant to dementia patients by manually review of 100 documents.
 ConclusionAutomated annotations must be validated. The platform we have developed allows efficient review and correction of annotations to allow models to be trained further or provide confidence that accuracy is sufficient for subsequent analysis. Implementation within our linked NCHA data platform will allow incorporation of text based data at scale.

ELaPro, a LOINC-mapped core dataset for top laboratory procedures of eligibility screening for clinical trials

BackgroundScreening for eligible patients continues to pose a great challenge for many clinical trials. This has led to a rapidly growing interest in standardizing computable representations of eligibility criteria (EC) in order to develop tools that leverage data from electronic health record (EHR) systems. Although laboratory procedures (LP) represent a common entity of EC that is readily available and retrievable from EHR systems, there is a lack of interoperable data models for this entity of EC. A public, specialized data model that utilizes international, widely-adopted terminology for LP, e.g. Logical Observation Identifiers Names and Codes (LOINC®), is much needed to support automated screening tools.ObjectiveThe aim of this study is to establish a core dataset for LP most frequently requested to recruit patients for clinical trials using LOINC terminology. Employing such a core dataset could enhance the interface between study feasibility platforms and EHR systems and significantly improve automatic patient recruitment.MethodsWe used a semi-automated approach to analyze 10,516 screening forms from the Medical Data Models (MDM) portal’s data repository that are pre-annotated with Unified Medical Language System (UMLS). An automated semantic analysis based on concept frequency is followed by an extensive manual expert review performed by physicians to analyze complex recruitment-relevant concepts not amenable to automatic approach.ResultsBased on analysis of 138,225 EC from 10,516 screening forms, 55 laboratory procedures represented 77.87% of all UMLS laboratory concept occurrences identified in the selected EC forms. We identified 26,413 unique UMLS concepts from 118 UMLS semantic types and covered the vast majority of Medical Subject Headings (MeSH) disease domains.ConclusionsOnly a small set of common LP covers the majority of laboratory concepts in screening EC forms which supports the feasibility of establishing a focused core dataset for LP. We present ELaPro, a novel, LOINC-mapped, core dataset for the most frequent 55 LP requested in screening for clinical trials. ELaPro is available in multiple machine-readable data formats like CSV, ODM and HL7 FHIR. The extensive manual curation of this large number of free-text EC as well as the combining of UMLS and LOINC terminologies distinguishes this specialized dataset from previous relevant datasets in the literature.

Abstract WP74: An Automated, Electronic Health Record-based Algorithm To Classify Ischemic Stroke Etiology

Introduction: Determining acute ischemic stroke (AIS) etiology is central to secondary stroke prevention, but can be diagnostically challenging. We built a stroke etiology classification model using electronic health record (EHR) data from non-cryptogenic AIS hospitalization encounters. Methods: We studied discharge summaries of non-cryptogenic AIS patients hospitalized at 2 academic, Comprehensive Stroke Centers from 2015-2020. The outcome was stroke etiology encoded in each institution’s stroke registry, abstracted from the discharge attending’s impression. We engineered variables including 1) clinical features from Unified Medical Language System concepts identified by a text mining tool, MetaMap , and 2) neuroimaging, echocardiographic, and laboratory features by regular expression operations. XGBoost classifier models were generated (80% training set, 20% testing set). A grid search of 36 model parameter value combinations and 10-fold cross validation were performed. We assessed variable importance in the optimized algorithm by calculating each feature’s Shapley value, the absolute value of the variable’s mean impact in predicting each specific stroke etiology. Results: The cohort was comprised of 1,993 patients (mean age 69 years, 55% male). We engineered a total of 2,004 features. Mean cross-validated accuracies of the best model in the training and validation sets were: 73.0% and 65.5%, respectively. Validation precision, recall, and F1 weighted scores were 66.2%, 65.5%, and 64.1% (mean cross-validated). The top 20 features contributing to etiology prediction are in Figure 1. Conclusion: We developed an EHR-based, automated tool to replicate a discharging attending’s stroke etiology diagnosis in non-cyptogenic patients. With further training in representative cohorts of non-cryptogenic and cryptogenic patients with an eventually diagnosed etiology, this algorithm may aide stroke etiology diagnosis before hospital discharge.

Text Mining of Adverse Events in Clinical Trials: Deep Learning Approach.

BackgroundPharmacovigilance and safety reporting, which involve processes for monitoring the use of medicines in clinical trials, play a critical role in the identification of previously unrecognized adverse events or changes in the patterns of adverse events.ObjectiveThis study aims to demonstrate the feasibility of automating the coding of adverse events described in the narrative section of the serious adverse event report forms to enable statistical analysis of the aforementioned patterns.MethodsWe used the Unified Medical Language System (UMLS) as the coding scheme, which integrates 217 source vocabularies, thus enabling coding against other relevant terminologies such as the International Classification of Diseases–10th Revision, Medical Dictionary for Regulatory Activities, and Systematized Nomenclature of Medicine). We used MetaMap, a highly configurable dictionary lookup software, to identify the mentions of the UMLS concepts. We trained a binary classifier using Bidirectional Encoder Representations from Transformers (BERT), a transformer-based language model that captures contextual relationships, to differentiate between mentions of the UMLS concepts that represented adverse events and those that did not.ResultsThe model achieved a high F1 score of 0.8080, despite the class imbalance. This is 10.15 percent points lower than human-like performance but also 17.45 percent points higher than that of the baseline approach.ConclusionsThese results confirmed that automated coding of adverse events described in the narrative section of serious adverse event reports is feasible. Once coded, adverse events can be statistically analyzed so that any correlations with the trialed medicines can be estimated in a timely fashion.

Rare Disease Identification from Clinical Notes with Ontologies and Weak Supervision.

The identification of rare diseases from clinical notes with Natural Language Processing (NLP) is challenging due to the few cases available for machine learning and the need of data annotation from clinical experts. We propose a method using ontologies and weak supervision. The approach includes two steps: (i) Text-to-UMLS, linking text mentions to concepts in Unified Medical Language System (UMLS), with a named entity linking tool (e.g. SemEHR) and weak supervision based on customised rules and Bidirectional Encoder Representations from Transformers (BERT) based contextual representations, and (ii) UMLS-to-ORDO, matching UMLS concepts to rare diseases in Orphanet Rare Disease Ontology (ORDO). Using MIMIC-III US intensive care discharge summaries as a case study, we show that the Text-to-UMLS process can be greatly improved with weak supervision, without any annotated data from domain experts. Our analysis shows that the overall pipeline processing discharge summaries can surface rare disease cases, which are mostly uncaptured in manual ICD codes of the hospital admissions.

A Silver Standard Biomedical Corpus for Arabic Language

The rapidly growing data in many areas, as well as in the biomedical domain, require the assistance of information extraction systems to acquire the much needed knowledge about specific entities such as proteins, drugs, or diseases practically within a short time. Annotated corpora serve the purpose of facilitating the process of building NLP systems. While colossal work has been done in this area for English language, other languages like Arabic seem to lack these resources, especially in the healthcare area. Therefore, in this work, we present a method to develop a silver standard medical corpus for the Arabic language with a dictionary as a minimal supervision tool. The corpus contains 49,856 sentences tagged with 13 entity types corresponding to a subset of UMLS (Unified Medical Language System) concept types. The evaluation of a subset of corpus showed the efficiency of the method used to annotate it with 90% accuracy.

Use of word and graph embedding to measure semantic relatedness between Unified Medical Language System concepts.

ObjectiveThe study sought to explore the use of deep learning techniques to measure the semantic relatedness between Unified Medical Language System (UMLS) concepts.Materials and MethodsConcept sentence embeddings were generated for UMLS concepts by applying the word embedding models BioWordVec and various flavors of BERT to concept sentences formed by concatenating UMLS terms. Graph embeddings were generated by the graph convolutional networks and 4 knowledge graph embedding models, using graphs built from UMLS hierarchical relations. Semantic relatedness was measured by the cosine between the concepts’ embedding vectors. Performance was compared with 2 traditional path-based (shortest path and Leacock-Chodorow) measurements and the publicly available concept embeddings, cui2vec, generated from large biomedical corpora. The concept sentence embeddings were also evaluated on a word sense disambiguation (WSD) task. Reference standards used included the semantic relatedness and semantic similarity datasets from the University of Minnesota, concept pairs generated from the Standardized MedDRA Queries and the MeSH (Medical Subject Headings) WSD corpus.ResultsSentence embeddings generated by BioWordVec outperformed all other methods used individually in semantic relatedness measurements. Graph convolutional network graph embedding uniformly outperformed path-based measurements and was better than some word embeddings for the Standardized MedDRA Queries dataset. When used together, combined word and graph embedding achieved the best performance in all datasets. For WSD, the enhanced versions of BERT outperformed BioWordVec.ConclusionsWord and graph embedding techniques can be used to harness terms and relations in the UMLS to measure semantic relatedness between concepts. Concept sentence embedding outperforms path-based measurements and cui2vec, and can be further enhanced by combining with graph embedding.

Assessing the enrichment of dietary supplement coverage in the Unified Medical Language System.

We sought to assess the need for additional coverage of dietary supplements (DS) in the Unified Medical Language System (UMLS) by investigating (1) the overlap between the integrated DIetary Supplements Knowledge base (iDISK) DS ingredient terminology and the UMLS and (2) the coverage of iDISK and the UMLS over DS mentions in the biomedical literature. We estimated the overlap between iDISK and the UMLS by mapping iDISK to the UMLS using exact and normalized strings. The coverage of iDISK and the UMLS over DS mentions in the biomedical literature was evaluated via a DS named-entity recognition (NER) task within PubMed abstracts. The coverage analysis revealed that only 30% of iDISK terms can be matched to the UMLS, although these cover over 99% of iDISK concepts. A manual review revealed that a majority of the unmatched terms represented new synonyms, rather than lexical variants. For NER, iDISK nearly doubles the precision and achieves a higher F1 score than the UMLS, while maintaining a competitive recall. While iDISK has significant concept overlap with the UMLS, it contains many novel synonyms. Furthermore, almost 3000 of these overlapping UMLS concepts are missing a DS designation, which could be provided by iDISK. The NER experiments show that the specialization of iDISK is useful for identifying DS mentions. Our results show that the DS representation in the UMLS could be enriched by adding DS designations to many concepts and by adding new synonyms.