1
|
Seinen TM, Kors JA, van Mulligen EM, Rijnbeek PR. Annotation-preserving machine translation of English corpora to validate Dutch clinical concept extraction tools. J Am Med Inform Assoc 2024:ocae159. [PMID: 38934643 DOI: 10.1093/jamia/ocae159] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/25/2024] [Revised: 05/24/2024] [Accepted: 06/10/2024] [Indexed: 06/28/2024] Open
Abstract
OBJECTIVE To explore the feasibility of validating Dutch concept extraction tools using annotated corpora translated from English, focusing on preserving annotations during translation and addressing the scarcity of non-English annotated clinical corpora. MATERIALS AND METHODS Three annotated corpora were standardized and translated from English to Dutch using 2 machine translation services, Google Translate and OpenAI GPT-4, with annotations preserved through a proposed method of embedding annotations in the text before translation. The performance of 2 concept extraction tools, MedSpaCy and MedCAT, was assessed across the corpora in both Dutch and English. RESULTS The translation process effectively generated Dutch annotated corpora and the concept extraction tools performed similarly in both English and Dutch. Although there were some differences in how annotations were preserved across translations, these did not affect extraction accuracy. Supervised MedCAT models consistently outperformed unsupervised models, whereas MedSpaCy demonstrated high recall but lower precision. DISCUSSION Our validation of Dutch concept extraction tools on corpora translated from English was successful, highlighting the efficacy of our annotation preservation method and the potential for efficiently creating multilingual corpora. Further improvements and comparisons of annotation preservation techniques and strategies for corpus synthesis could lead to more efficient development of multilingual corpora and accurate non-English concept extraction tools. CONCLUSION This study has demonstrated that translated English corpora can be used to validate non-English concept extraction tools. The annotation preservation method used during translation proved effective, and future research can apply this corpus translation method to additional languages and clinical settings.
Collapse
Affiliation(s)
- Tom M Seinen
- Department of Medical Informatics, Erasmus University Medical Center, 3015 GD Rotterdam, The Netherlands
| | - Jan A Kors
- Department of Medical Informatics, Erasmus University Medical Center, 3015 GD Rotterdam, The Netherlands
| | - Erik M van Mulligen
- Department of Medical Informatics, Erasmus University Medical Center, 3015 GD Rotterdam, The Netherlands
| | - Peter R Rijnbeek
- Department of Medical Informatics, Erasmus University Medical Center, 3015 GD Rotterdam, The Netherlands
| |
Collapse
|
2
|
Moreno AC, Bitterman DS. Toward Clinical-Grade Evaluation of Large Language Models. Int J Radiat Oncol Biol Phys 2024; 118:916-920. [PMID: 38401979 PMCID: PMC11221761 DOI: 10.1016/j.ijrobp.2023.11.012] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/30/2023] [Accepted: 11/05/2023] [Indexed: 02/26/2024]
Affiliation(s)
- Amy C Moreno
- Department of Radiation Oncology, University of Texas MD Anderson Cancer Center, Houston, Texas
| | - Danielle S Bitterman
- Department of Radiation Oncology, Brigham and Women's Hospital/Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts; Artificial Intelligence in Medicine Program, Mass General Brigham, Harvard Medical School, Boston, Massachusetts.
| |
Collapse
|
3
|
Irrera O, Marchesin S, Silvello G. MetaTron: advancing biomedical annotation empowering relation annotation and collaboration. BMC Bioinformatics 2024; 25:112. [PMID: 38486137 PMCID: PMC10941452 DOI: 10.1186/s12859-024-05730-9] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/26/2023] [Accepted: 03/04/2024] [Indexed: 03/17/2024] Open
Abstract
BACKGROUND The constant growth of biomedical data is accompanied by the need for new methodologies to effectively and efficiently extract machine-readable knowledge for training and testing purposes. A crucial aspect in this regard is creating large, often manually or semi-manually, annotated corpora vital for developing effective and efficient methods for tasks like relation extraction, topic recognition, and entity linking. However, manual annotation is expensive and time-consuming especially if not assisted by interactive, intuitive, and collaborative computer-aided tools. To support healthcare experts in the annotation process and foster annotated corpora creation, we present MetaTron. MetaTron is an open-source and free-to-use web-based annotation tool to annotate biomedical data interactively and collaboratively; it supports both mention-level and document-level annotations also integrating automatic built-in predictions. Moreover, MetaTron enables relation annotation with the support of ontologies, functionalities often overlooked by off-the-shelf annotation tools. RESULTS We conducted a qualitative analysis to compare MetaTron with a set of manual annotation tools including TeamTat, INCEpTION, LightTag, MedTAG, and brat, on three sets of criteria: technical, data, and functional. A quantitative evaluation allowed us to assess MetaTron performances in terms of time and number of clicks to annotate a set of documents. The results indicated that MetaTron fulfills almost all the selected criteria and achieves the best performances. CONCLUSIONS MetaTron stands out as one of the few annotation tools targeting the biomedical domain supporting the annotation of relations, and fully customizable with documents in several formats-PDF included, as well as abstracts retrieved from PubMed, Semantic Scholar, and OpenAIRE. To meet any user need, we released MetaTron both as an online instance and as a Docker image locally deployable.
Collapse
Affiliation(s)
- Ornella Irrera
- Department of Information Engineering, University of Padova, Padua, Italy.
| | - Stefano Marchesin
- Department of Information Engineering, University of Padova, Padua, Italy
| | - Gianmaria Silvello
- Department of Information Engineering, University of Padova, Padua, Italy
| |
Collapse
|
4
|
Simoulin A, Thiebaut N, Neuberger K, Ibnouhsein I, Brunel N, Viné R, Bousquet N, Latapy J, Reix N, Molière S, Lodi M, Mathelin C. From free-text electronic health records to structured cohorts: Onconum, an innovative methodology for real-world data mining in breast cancer. COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE 2023; 240:107693. [PMID: 37453367 DOI: 10.1016/j.cmpb.2023.107693] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 05/30/2022] [Revised: 05/25/2023] [Accepted: 06/23/2023] [Indexed: 07/18/2023]
Abstract
PURPOSE A considerable amount of valuable information is present in electronic health records (EHRs) however it remains inaccessible because it is embedded into unstructured narrative documents that cannot be easily analyzed. We wanted to develop and evaluate a methodology able to extract and structure information from electronic health records in breast cancer. METHODS We developed a software platform called Onconum (ClinicalTrials.gov Identifier: NCT02810093) which uses a hybrid method relying on machine learning approaches and rule-based lexical methods. It is based on natural language processing techniques that allows a targeted analysis of free-text medical data related to breast cancer, independently of any pre-existing dictionary, in a French context (available in N files). We then evaluated it on a validation cohort called Senometry. FINDINGS Senometry cohort included 9,599 patients with breast cancer (both invasive and in situ), treated between 2000 and 2017 in the breast cancer unit of Strasbourg University Hospitals. Extraction rates ranged from 45 to 100%, depending on the type of each parameter. Precision of extracted information was 68%-94% compared to a structured cohort, and 89%-98% compared to manually structured databases and it retrieved more rare occurrences compared to another database search engine (+17%). INTERPRETATION This innovative method can accurately structure relevant medical information embedded in EHRs in the context of breast cancer. Missing data handling is the main limitation of this method however multiple sources can be incorporated to reduce this limit. Nevertheless, this methodology does not need neither pre-existing dictionaries nor manually annotated corpora. It can therefore be easily implemented in non-English-speaking countries and in other diseases outside breast cancer, and it allows prospective inclusion of new patients.
Collapse
Affiliation(s)
| | | | | | | | | | | | - Nicolas Bousquet
- Quantmetry, 52 rue d'Anjou, 75008 Paris, France; Sorbonne University, 4 place Jussieu, 75005 Paris, France
| | | | - Nathalie Reix
- ICube UMR 7537, Strasbourg University / CNRS, Fédération de Médecine Translationnelle de Strasbourg, 67200 Strasbourg, France; Biochemistry and Molecular Biology Laboratory, Strasbourg University Hospitals, 1 place de l'Hôpital, 67091 Strasbourg, France
| | - Sébastien Molière
- Radiology Department, Strasbourg University Hospitals, 1 avenue Molière, 67098 Strasbourg, France
| | - Massimo Lodi
- Institut de cancérologie Strasbourg Europe (ICANS), 17 avenue Albert Calmette, 67033 Strasbourg Cedex, France; Department of Functional Genomics and Cancer, Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS UMR 7104, INSERM U964, Strasbourg University, Illkirch, France; Strasbourg University Hospitals, 1 place de l'Hôpital, 67091 Strasbourg, France.
| | - Carole Mathelin
- Institut de cancérologie Strasbourg Europe (ICANS), 17 avenue Albert Calmette, 67033 Strasbourg Cedex, France; Department of Functional Genomics and Cancer, Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS UMR 7104, INSERM U964, Strasbourg University, Illkirch, France; Strasbourg University Hospitals, 1 place de l'Hôpital, 67091 Strasbourg, France.
| |
Collapse
|
5
|
Fraile Navarro D, Ijaz K, Rezazadegan D, Rahimi-Ardabili H, Dras M, Coiera E, Berkovsky S. Clinical named entity recognition and relation extraction using natural language processing of medical free text: A systematic review. Int J Med Inform 2023; 177:105122. [PMID: 37295138 DOI: 10.1016/j.ijmedinf.2023.105122] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/09/2022] [Revised: 04/14/2023] [Accepted: 06/03/2023] [Indexed: 06/12/2023]
Abstract
BACKGROUND Natural Language Processing (NLP) applications have developed over the past years in various fields including its application to clinical free text for named entity recognition and relation extraction. However, there has been rapid developments the last few years that there's currently no overview of it. Moreover, it is unclear how these models and tools have been translated into clinical practice. We aim to synthesize and review these developments. METHODS We reviewed literature from 2010 to date, searching PubMed, Scopus, the Association of Computational Linguistics (ACL), and Association of Computer Machinery (ACM) libraries for studies of NLP systems performing general-purpose (i.e., not disease- or treatment-specific) information extraction and relation extraction tasks in unstructured clinical text (e.g., discharge summaries). RESULTS We included in the review 94 studies with 30 studies published in the last three years. Machine learning methods were used in 68 studies, rule-based in 5 studies, and both in 22 studies. 63 studies focused on Named Entity Recognition, 13 on Relation Extraction and 18 performed both. The most frequently extracted entities were "problem", "test" and "treatment". 72 studies used public datasets and 22 studies used proprietary datasets alone. Only 14 studies defined clearly a clinical or information task to be addressed by the system and just three studies reported its use outside the experimental setting. Only 7 studies shared a pre-trained model and only 8 an available software tool. DISCUSSION Machine learning-based methods have dominated the NLP field on information extraction tasks. More recently, Transformer-based language models are taking the lead and showing the strongest performance. However, these developments are mostly based on a few datasets and generic annotations, with very few real-world use cases. This may raise questions about the generalizability of findings, translation into practice and highlights the need for robust clinical evaluation.
Collapse
Affiliation(s)
- David Fraile Navarro
- Centre for Health Informatics, Australian Institute of Health Innovation, Macquarie University, Sydney, Australia.
| | - Kiran Ijaz
- Centre for Health Informatics, Australian Institute of Health Innovation, Macquarie University, Sydney, Australia
| | - Dana Rezazadegan
- Department of Computer Science and Software Engineering. School of Software and Electrical Engineering, Swinburne University of Technology, Melbourne, Australia
| | - Hania Rahimi-Ardabili
- Centre for Health Informatics, Australian Institute of Health Innovation, Macquarie University, Sydney, Australia
| | - Mark Dras
- Department of Computing, Macquarie University, Sydney, Australia
| | - Enrico Coiera
- Centre for Health Informatics, Australian Institute of Health Innovation, Macquarie University, Sydney, Australia
| | - Shlomo Berkovsky
- Centre for Health Informatics, Australian Institute of Health Innovation, Macquarie University, Sydney, Australia
| |
Collapse
|
6
|
Ismail A, Al-Zoubi T, El Naqa I, Saeed H. The role of artificial intelligence in hastening time to recruitment in clinical trials. BJR Open 2023; 5:20220023. [PMID: 37953865 PMCID: PMC10636341 DOI: 10.1259/bjro.20220023] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/28/2022] [Revised: 03/20/2023] [Accepted: 04/11/2023] [Indexed: 09/01/2023] Open
Abstract
Novel and developing artificial intelligence (AI) systems can be integrated into healthcare settings in numerous ways. For example, in the case of automated image classification and natural language processing, AI systems are beginning to demonstrate near expert level performance in detecting abnormalities such as seizure activity. This paper, however, focuses on AI integration into clinical trials. During the clinical trial recruitment process, considerable labor and time is spent sifting through electronic health record and interviewing patients. With the advancement of deep learning techniques such as natural language processing, intricate electronic health record data can be efficiently processed. This provides utility to workflows such as recruitment for clinical trials. Studies are starting to show promise in shortening the time to recruitment and reducing workload for those involved in clinical trial design. Additionally, numerous guidelines are being constructed to encourage integration of AI into the healthcare setting with meaningful impact. The goal would be to improve the clinical trial process by reducing bias in patient composition, improving retention of participants, and lowering costs and labor.
Collapse
Affiliation(s)
- Abdalah Ismail
- Advocate Aurora Health Care Department of Diagnostic Radiology, Aurora, United States
| | | | | | - Hina Saeed
- Lynn Cancer Institute-Baptist Health City, Boca Raton, United States
| |
Collapse
|
7
|
Eikelboom WS, Singleton EH, van den Berg E, de Boer C, Coesmans M, Goudzwaard JA, Vijverberg EGB, Pan M, Gouw C, Mol MO, Gillissen F, Fieldhouse JLP, Pijnenburg YAL, van der Flier WM, van Swieten JC, Ossenkoppele R, Kors JA, Papma JM. The reporting of neuropsychiatric symptoms in electronic health records of individuals with Alzheimer's disease: a natural language processing study. Alzheimers Res Ther 2023; 15:94. [PMID: 37173801 PMCID: PMC10176879 DOI: 10.1186/s13195-023-01240-7] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/20/2022] [Accepted: 05/05/2023] [Indexed: 05/15/2023]
Abstract
BACKGROUND Neuropsychiatric symptoms (NPS) are prevalent in the early clinical stages of Alzheimer's disease (AD) according to proxy-based instruments. Little is known about which NPS clinicians report and whether their judgment aligns with proxy-based instruments. We used natural language processing (NLP) to classify NPS in electronic health records (EHRs) to estimate the reporting of NPS in symptomatic AD at the memory clinic according to clinicians. Next, we compared NPS as reported in EHRs and NPS reported by caregivers on the Neuropsychiatric Inventory (NPI). METHODS Two academic memory clinic cohorts were used: the Amsterdam UMC (n = 3001) and the Erasmus MC (n = 646). Patients included in these cohorts had MCI, AD dementia, or mixed AD/VaD dementia. Ten trained clinicians annotated 13 types of NPS in a randomly selected training set of n = 500 EHRs from the Amsterdam UMC cohort and in a test set of n = 250 EHRs from the Erasmus MC cohort. For each NPS, a generalized linear classifier was trained and internally and externally validated. Prevalence estimates of NPS were adjusted for the imperfect sensitivity and specificity of each classifier. Intra-individual comparison of the NPS classified in EHRs and NPS reported on the NPI were conducted in a subsample (59%). RESULTS Internal validation performance of the classifiers was excellent (AUC range: 0.81-0.91), but external validation performance decreased (AUC range: 0.51-0.93). NPS were prevalent in EHRs from the Amsterdam UMC, especially apathy (adjusted prevalence = 69.4%), anxiety (adjusted prevalence = 53.7%), aberrant motor behavior (adjusted prevalence = 47.5%), irritability (adjusted prevalence = 42.6%), and depression (adjusted prevalence = 38.5%). The ranking of NPS was similar for EHRs from the Erasmus MC, although not all classifiers obtained valid prevalence estimates due to low specificity. In both cohorts, there was minimal agreement between NPS classified in the EHRs and NPS reported on the NPI (all kappa coefficients < 0.28), with substantially more reports of NPS in EHRs than on NPI assessments. CONCLUSIONS NLP classifiers performed well in detecting a wide range of NPS in EHRs of patients with symptomatic AD visiting the memory clinic and showed that clinicians frequently reported NPS in these EHRs. Clinicians generally reported more NPS in EHRs than caregivers reported on the NPI.
Collapse
Affiliation(s)
- Willem S Eikelboom
- Department of Neurology and Alzheimer Center Erasmus MC, Erasmus MC University Medical Center, PO Box 2040, 3000 CA, Rotterdam, the Netherlands.
| | - Ellen H Singleton
- Department of Neurology, Alzheimer Center Amsterdam, Amsterdam University Medical Centers, Amsterdam, the Netherlands
| | - Esther van den Berg
- Department of Neurology and Alzheimer Center Erasmus MC, Erasmus MC University Medical Center, PO Box 2040, 3000 CA, Rotterdam, the Netherlands
| | - Casper de Boer
- Department of Neurology, Alzheimer Center Amsterdam, Amsterdam University Medical Centers, Amsterdam, the Netherlands
| | - Michiel Coesmans
- Department of Psychiatry, Erasmus MC University Medical Center, Rotterdam, the Netherlands
| | - Jeannette A Goudzwaard
- Department of Internal Medicine, Section of Geriatrics, Erasmus MC University Medical Center, Rotterdam, the Netherlands
| | - Everard G B Vijverberg
- Department of Neurology, Alzheimer Center Amsterdam, Amsterdam University Medical Centers, Amsterdam, the Netherlands
| | - Michel Pan
- Department of Neurology and Alzheimer Center Erasmus MC, Erasmus MC University Medical Center, PO Box 2040, 3000 CA, Rotterdam, the Netherlands
| | - Cornalijn Gouw
- Department of Psychiatry, Erasmus MC University Medical Center, Rotterdam, the Netherlands
| | - Merel O Mol
- Department of Neurology and Alzheimer Center Erasmus MC, Erasmus MC University Medical Center, PO Box 2040, 3000 CA, Rotterdam, the Netherlands
| | - Freek Gillissen
- Department of Neurology, Alzheimer Center Amsterdam, Amsterdam University Medical Centers, Amsterdam, the Netherlands
| | - Jay L P Fieldhouse
- Department of Neurology, Alzheimer Center Amsterdam, Amsterdam University Medical Centers, Amsterdam, the Netherlands
| | - Yolande A L Pijnenburg
- Department of Neurology, Alzheimer Center Amsterdam, Amsterdam University Medical Centers, Amsterdam, the Netherlands
| | - Wiesje M van der Flier
- Department of Neurology, Alzheimer Center Amsterdam, Amsterdam University Medical Centers, Amsterdam, the Netherlands
| | - John C van Swieten
- Department of Neurology and Alzheimer Center Erasmus MC, Erasmus MC University Medical Center, PO Box 2040, 3000 CA, Rotterdam, the Netherlands
| | - Rik Ossenkoppele
- Department of Neurology, Alzheimer Center Amsterdam, Amsterdam University Medical Centers, Amsterdam, the Netherlands
- Clinical Memory Research Unit, Lund University, Malmö, Sweden
| | - Jan A Kors
- Department of Medical Informatics, Erasmus MC University Medical Center, Rotterdam, the Netherlands
| | - Janne M Papma
- Department of Neurology and Alzheimer Center Erasmus MC, Erasmus MC University Medical Center, PO Box 2040, 3000 CA, Rotterdam, the Netherlands
| |
Collapse
|
8
|
Dong H, Suárez-Paniagua V, Zhang H, Wang M, Casey A, Davidson E, Chen J, Alex B, Whiteley W, Wu H. Ontology-driven and weakly supervised rare disease identification from clinical notes. BMC Med Inform Decis Mak 2023; 23:86. [PMID: 37147628 PMCID: PMC10162001 DOI: 10.1186/s12911-023-02181-9] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/09/2022] [Accepted: 04/21/2023] [Indexed: 05/07/2023] Open
Abstract
BACKGROUND Computational 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. METHODS We 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. RESULTS The 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). CONCLUSION The 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.
Collapse
Affiliation(s)
- Hang Dong
- Centre for Medical Informatics, Usher Institute of Population Health Sciences and Informatics, University of Edinburgh, Edinburgh, United Kingdom.
- Health Data Research UK, London, United Kingdom.
- Department of Computer Science, University of Oxford, Oxford, United Kingdom.
| | - Víctor Suárez-Paniagua
- Centre for Medical Informatics, Usher Institute of Population Health Sciences and Informatics, University of Edinburgh, Edinburgh, United Kingdom
- Health Data Research UK, London, United Kingdom
| | - Huayu Zhang
- Advanced Care Research Centre, Usher Institute, University of Edinburgh, Edinburgh, United Kingdom
| | - Minhong Wang
- Institute of Health Informatics, University College London, London, United Kingdom
| | - Arlene Casey
- Advanced Care Research Centre, Usher Institute, University of Edinburgh, Edinburgh, United Kingdom
| | - Emma Davidson
- Centre for Clinical Brain Sciences, University of Edinburgh, Edinburgh, United Kingdom
| | - Jiaoyan Chen
- Department of Computer Science, The University of Manchester, Manchester, United Kingdom
| | - Beatrice Alex
- Edinburgh Futures Institute, University of Edinburgh, Edinburgh, United Kingdom
| | - William Whiteley
- Health Data Research UK, London, United Kingdom
- Centre for Clinical Brain Sciences, University of Edinburgh, Edinburgh, United Kingdom
| | - Honghan Wu
- Health Data Research UK, London, United Kingdom.
- Institute of Health Informatics, University College London, London, United Kingdom.
| |
Collapse
|
9
|
Graeßner M, Jungwirth B, Frank E, Schaller SJ, Kochs E, Ulm K, Blobner M, Ulm B, Podtschaske AH, Kagerbauer SM. Enabling personalized perioperative risk prediction by using a machine-learning model based on preoperative data. Sci Rep 2023; 13:7128. [PMID: 37130884 PMCID: PMC10153050 DOI: 10.1038/s41598-023-33981-8] [Citation(s) in RCA: 4] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Grants] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/21/2022] [Accepted: 04/21/2023] [Indexed: 05/04/2023] Open
Abstract
Preoperative risk assessment is essential for shared decision-making and adequate perioperative care. Common scores provide limited predictive quality and lack personalized information. The aim of this study was to create an interpretable machine-learning-based model to assess the patient's individual risk of postoperative mortality based on preoperative data to allow analysis of personal risk factors. After ethical approval, a model for prediction of postoperative in-hospital mortality based on preoperative data of 66,846 patients undergoing elective non-cardiac surgery between June 2014 and March 2020 was created with extreme gradient boosting. Model performance and the most relevant parameters were shown using receiver operating characteristic (ROC-) and precision-recall (PR-) curves and importance plots. Individual risks of index patients were presented in waterfall diagrams. The model included 201 features and showed good predictive abilities with an area under receiver operating characteristic (AUROC) curve of 0.95 and an area under precision-recall curve (AUPRC) of 0.109. The feature with the highest information gain was the preoperative order for red packed cell concentrates followed by age and c-reactive protein. Individual risk factors could be identified on patient level. We created a highly accurate and interpretable machine learning model to preoperatively predict the risk of postoperative in-hospital mortality. The algorithm can be used to identify factors susceptible to preoperative optimization measures and to identify risk factors influencing individual patient risk.
Collapse
Affiliation(s)
- Martin Graeßner
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, Technical University of Munich, Munich, Germany
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, University Hospital Ulm, Albert-Einstein-Allee 23, 89081, Ulm, Germany
| | - Bettina Jungwirth
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, Technical University of Munich, Munich, Germany
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, University Hospital Ulm, Albert-Einstein-Allee 23, 89081, Ulm, Germany
| | - Elke Frank
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, University Hospital Ulm, Albert-Einstein-Allee 23, 89081, Ulm, Germany
- Commercial department, Klinikum rechts der isar, Technical University of Munich, Munich, Germany
| | - Stefan Josef Schaller
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, Technical University of Munich, Munich, Germany
- Department of Anaesthesiology and Operative Intensive Care Medicine (CVK, CCM), Charité-Universitätsmedizin Berlin, Corporate Member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Berlin, Germany
| | - Eberhard Kochs
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, Technical University of Munich, Munich, Germany
| | - Kurt Ulm
- Department of Medical Statistics and Epidemiology, School of Medicine, Technical University of Munich, Munich, Germany
| | - Manfred Blobner
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, Technical University of Munich, Munich, Germany
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, University Hospital Ulm, Albert-Einstein-Allee 23, 89081, Ulm, Germany
| | - Bernhard Ulm
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, Technical University of Munich, Munich, Germany
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, University Hospital Ulm, Albert-Einstein-Allee 23, 89081, Ulm, Germany
| | - Armin Horst Podtschaske
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, Technical University of Munich, Munich, Germany
| | - Simone Maria Kagerbauer
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, Technical University of Munich, Munich, Germany.
- Department of Anaesthesiology and Intensive Care Medicine, School of Medicine, University Hospital Ulm, Albert-Einstein-Allee 23, 89081, Ulm, Germany.
| |
Collapse
|
10
|
Pethani F, Dunn AG. Natural language processing for clinical notes in dentistry: A systematic review. J Biomed Inform 2023; 138:104282. [PMID: 36623780 DOI: 10.1016/j.jbi.2023.104282] [Citation(s) in RCA: 4] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/29/2022] [Revised: 12/01/2022] [Accepted: 01/04/2023] [Indexed: 01/09/2023]
Abstract
OBJECTIVE To identify and synthesise research on applications of natural language processing (NLP) for information extraction and retrieval from clinical notes in dentistry. MATERIALS AND METHODS A predefined search strategy was applied in EMBASE, CINAHL and Medline. Studies eligible for inclusion were those that that described, evaluated, or applied NLP to clinical notes containing either human or simulated patient information. Quality of the study design and reporting was independently assessed based on a set of questions derived from relevant tools including CHecklist for critical Appraisal and data extraction for systematic Reviews of prediction Modelling Studies (CHARMS). A narrative synthesis was conducted to present the results. RESULTS Of the 17 included studies, 10 developed and evaluated NLP methods and 7 described applications of NLP-based information retrieval methods in dental records. Studies were published between 2015 and 2021, most were missing key details needed for reproducibility, and there was no consistency in design or reporting. The 10 studies developing or evaluating NLP methods used document classification or entity extraction, and 4 compared NLP methods to non-NLP methods. The quality of reporting on NLP studies in dentistry has modestly improved over time. CONCLUSIONS Study design heterogeneity and incomplete reporting of studies currently limits our ability to synthesise NLP applications in dental records. Standardisation of reporting and improved connections between NLP methods and applied NLP in dentistry may improve how we can make use of clinical notes from dentistry in population health or decision support systems. PROTOCOL REGISTRATION PROSPERO CRD42021227823.
Collapse
Affiliation(s)
- Farhana Pethani
- Biomedical Informatics and Digital Health, Faculty of Medicine and Health, the University of Sydney, Sydney, Australia
| | - Adam G Dunn
- Biomedical Informatics and Digital Health, Faculty of Medicine and Health, the University of Sydney, Sydney, Australia.
| |
Collapse
|
11
|
Zhao Y, Howard R, Amorrortu RP, Stewart SC, Wang X, Calip GS, Rollison DE. Assessing the Contribution of Scanned Outside Documents to the Completeness of Real-World Data Abstraction. JCO Clin Cancer Inform 2023; 7:e2200118. [PMID: 36791386 DOI: 10.1200/cci.22.00118] [Citation(s) in RCA: 2] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 02/17/2023] Open
Abstract
PURPOSE Electronic health record (EHR) data are widely used in precision medicine, quality improvement, disease surveillance, and population health management. However, a significant amount of EHR data are stored in unstructured formats including scanned documents external to the treatment facility presenting an informatics challenge for secondary use. Studies are needed to characterize the clinical information uniquely available in scanned outside documents (SODs) to understand to what extent the availability of such information affects the use of these real-world data for cancer research. MATERIALS AND METHODS Two independent EHR data abstractions capturing 30 variables commonly used in oncology research were conducted for 125 patients treated for advanced non-small-cell lung cancer at a comprehensive cancer center, with and without consideration of SODs. Completeness and concordance were compared between the two abstractions, overall, and by patient groups and variable types. RESULTS The overall completeness of the data with SODs was 77.6% as compared with 54.3% for the abstraction without SODs. The differences in completeness were driven by data related to biomarker tests, which were more likely to be uniquely available in SODs. Such data were prone to missingness among patients who were diagnosed externally. CONCLUSION There were no major differences in completeness between the two abstractions by demographics, diagnosis, disease progression, performance status, or oral therapy use. However, biomarker data were more likely to be uniquely contained in the SODs. Our findings may help cancer centers prioritize the types of SOD data being abstracted for research or other secondary purposes.
Collapse
Affiliation(s)
- Yayi Zhao
- Department of Cancer Epidemiology, Moffitt Cancer Center, Tampa, FL
| | - Rachel Howard
- Department of Health Informatics, Moffitt Cancer Center, Tampa, FL
| | | | | | | | - Gregory S Calip
- Flatiron Health, Inc., New York, NY.,University of Illinois Chicago, Center for Pharmacoepidemiology and Pharmacoeconomic Research, Chicago, IL
| | - Dana E Rollison
- Department of Cancer Epidemiology, Moffitt Cancer Center, Tampa, FL
| |
Collapse
|
12
|
Adverse drug event detection using natural language processing: A scoping review of supervised learning methods. PLoS One 2023; 18:e0279842. [PMID: 36595517 PMCID: PMC9810201 DOI: 10.1371/journal.pone.0279842] [Citation(s) in RCA: 7] [Impact Index Per Article: 7.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/22/2022] [Accepted: 12/15/2022] [Indexed: 01/04/2023] Open
Abstract
To reduce adverse drug events (ADEs), hospitals need a system to support them in monitoring ADE occurrence routinely, rapidly, and at scale. Natural language processing (NLP), a computerized approach to analyze text data, has shown promising results for the purpose of ADE detection in the context of pharmacovigilance. However, a detailed qualitative assessment and critical appraisal of NLP methods for ADE detection in the context of ADE monitoring in hospitals is lacking. Therefore, we have conducted a scoping review to close this knowledge gap, and to provide directions for future research and practice. We included articles where NLP was applied to detect ADEs in clinical narratives within electronic health records of inpatients. Quantitative and qualitative data items relating to NLP methods were extracted and critically appraised. Out of 1,065 articles screened for eligibility, 29 articles met the inclusion criteria. Most frequent tasks included named entity recognition (n = 17; 58.6%) and relation extraction/classification (n = 15; 51.7%). Clinical involvement was reported in nine studies (31%). Multiple NLP modelling approaches seem suitable, with Long Short Term Memory and Conditional Random Field methods most commonly used. Although reported overall performance of the systems was high, it provides an inflated impression given a steep drop in performance when predicting the ADE entity or ADE relation class. When annotating corpora, treating an ADE as a relation between a drug and non-drug entity seems the best practice. Future research should focus on semi-automated methods to reduce the manual annotation effort, and examine implementation of the NLP methods in practice.
Collapse
|
13
|
Carrillo-Larco RM, Castillo-Cara M, Lovón-Melgarejo J. Government plans in the 2016 and 2021 Peruvian presidential elections: A natural language processing analysis of the health chapters. Wellcome Open Res 2022. [DOI: 10.12688/wellcomeopenres.16867.5] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
Background: While clinical medicine has exploded, electronic health records for Natural Language Processing (NLP) analyses, public health, and health policy research have not yet adopted these algorithms. We aimed to dissect the health chapters of the government plans of the 2016 and 2021 Peruvian presidential elections, and to compare different NLP algorithms. Methods: From the government plans (18 in 2016; 19 in 2021) we extracted each sentence from the health chapters. We used five NLP algorithms to extract keywords and phrases from each plan: Term Frequency–Inverse Document Frequency (TF-IDF), Latent Dirichlet Allocation (LDA), TextRank, Keywords Bidirectional Encoder Representations from Transformers (KeyBERT), and Rapid Automatic Keywords Extraction (Rake). Results: In 2016 we analysed 630 sentences, whereas in 2021 there were 1,685 sentences. The TF-IDF algorithm showed that in 2016, 26 terms appeared with a frequency of 0.08 or greater, while in 2021 27 terms met this criterion. The LDA algorithm defined two groups. The first included terms related to things the population would receive (e.g., ’insurance’), while the second included terms about the health system (e.g., ’capacity’). In 2021, most of the government plans belonged to the second group. The TextRank analysis provided keywords showing that ’universal health coverage’ appeared frequently in 2016, while in 2021 keywords about the COVID-19 pandemic were often found. The KeyBERT algorithm provided keywords based on the context of the text. These keywords identified some underlying characteristics of the political party (e.g., political spectrum such as left-wing). The Rake algorithm delivered phrases, in which we found ’universal health coverage’ in 2016 and 2021. Conclusion: The NLP analysis could be used to inform on the underlying priorities in each government plan. NLP analysis could also be included in research of health policies and politics during general elections and provide informative summaries for the general population.
Collapse
|
14
|
Pavel A, Saarimäki LA, Möbus L, Federico A, Serra A, Greco D. The potential of a data centred approach & knowledge graph data representation in chemical safety and drug design. Comput Struct Biotechnol J 2022; 20:4837-4849. [PMID: 36147662 PMCID: PMC9464643 DOI: 10.1016/j.csbj.2022.08.061] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/29/2022] [Revised: 08/26/2022] [Accepted: 08/26/2022] [Indexed: 11/20/2022] Open
Abstract
Big Data pervades nearly all areas of life sciences, yet the analysis of large integrated data sets remains a major challenge. Moreover, the field of life sciences is highly fragmented and, consequently, so is its data, knowledge, and standards. This, in turn, makes integrated data analysis and knowledge gathering across sub-fields a demanding task. At the same time, the integration of various research angles and data types is crucial for modelling the complexity of organisms and biological processes in a holistic manner. This is especially valid in the context of drug development and chemical safety assessment where computational methods can provide solutions for the urgent need of fast, effective, and sustainable approaches. At the same time, such computational methods require the development of methodologies suitable for an integrated and data centred Big Data view. Here we discuss Knowledge Graphs (KG) as a solution to a data centred analysis approach for drug and chemical development and safety assessment. KGs are knowledge bases, data analysis engines, and knowledge discovery systems all in one, allowing them to be used from simple data retrieval, over meta-analysis to complex predictive and knowledge discovery systems. Therefore, KGs have immense potential to advance the data centred approach, the re-usability, and informativity of data. Furthermore, they can improve the power of analysis, and the complexity of modelled processes, all while providing knowledge in a natively human understandable network data model.
Collapse
Affiliation(s)
- Alisa Pavel
- Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland.,BioMediTech Institute, Tampere University, Tampere, Finland.,Finnish Hub for Development and Validation of Integrated Approaches (FHAIVE), Tampere, Finland
| | - Laura A Saarimäki
- Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland.,BioMediTech Institute, Tampere University, Tampere, Finland.,Finnish Hub for Development and Validation of Integrated Approaches (FHAIVE), Tampere, Finland
| | - Lena Möbus
- Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland.,BioMediTech Institute, Tampere University, Tampere, Finland.,Finnish Hub for Development and Validation of Integrated Approaches (FHAIVE), Tampere, Finland
| | - Antonio Federico
- Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland.,BioMediTech Institute, Tampere University, Tampere, Finland.,Finnish Hub for Development and Validation of Integrated Approaches (FHAIVE), Tampere, Finland
| | - Angela Serra
- Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland.,BioMediTech Institute, Tampere University, Tampere, Finland.,Finnish Hub for Development and Validation of Integrated Approaches (FHAIVE), Tampere, Finland
| | - Dario Greco
- Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland.,BioMediTech Institute, Tampere University, Tampere, Finland.,Finnish Hub for Development and Validation of Integrated Approaches (FHAIVE), Tampere, Finland.,Institute of Biotechnology, University of Helsinki, Helsinki, Finland
| |
Collapse
|
15
|
Karhade AV, Oosterhoff JHF, Groot OQ, Agaronnik N, Ehresman J, Bongers MER, Jaarsma RL, Poonnoose SI, Sciubba DM, Tobert DG, Doornberg JN, Schwab JH. Can We Geographically Validate a Natural Language Processing Algorithm for Automated Detection of Incidental Durotomy Across Three Independent Cohorts From Two Continents? Clin Orthop Relat Res 2022; 480:1766-1775. [PMID: 35412473 PMCID: PMC9384904 DOI: 10.1097/corr.0000000000002200] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 11/24/2021] [Accepted: 03/11/2022] [Indexed: 01/31/2023]
Abstract
BACKGROUND Incidental durotomy is an intraoperative complication in spine surgery that can lead to postoperative complications, increased length of stay, and higher healthcare costs. Natural language processing (NLP) is an artificial intelligence method that assists in understanding free-text notes that may be useful in the automated surveillance of adverse events in orthopaedic surgery. A previously developed NLP algorithm is highly accurate in the detection of incidental durotomy on internal validation and external validation in an independent cohort from the same country. External validation in a cohort with linguistic differences is required to assess the transportability of the developed algorithm, referred to geographical validation. Ideally, the performance of a prediction model, the NLP algorithm, is constant across geographic regions to ensure reproducibility and model validity. QUESTION/PURPOSE Can we geographically validate an NLP algorithm for the automated detection of incidental durotomy across three independent cohorts from two continents? METHODS Patients 18 years or older undergoing a primary procedure of (thoraco)lumbar spine surgery were included. In Massachusetts, between January 2000 and June 2018, 1000 patients were included from two academic and three community medical centers. In Maryland, between July 2016 and November 2018, 1279 patients were included from one academic center, and in Australia, between January 2010 and December 2019, 944 patients were included from one academic center. The authors retrospectively studied the free-text operative notes of included patients for the primary outcome that was defined as intraoperative durotomy. Incidental durotomy occurred in 9% (93 of 1000), 8% (108 of 1279), and 6% (58 of 944) of the patients, respectively, in the Massachusetts, Maryland, and Australia cohorts. No missing reports were observed. Three datasets (Massachusetts, Australian, and combined Massachusetts and Australian) were divided into training and holdout test sets in an 80:20 ratio. An extreme gradient boosting (an efficient and flexible tree-based algorithm) NLP algorithm was individually trained on each training set, and the performance of the three NLP algorithms (respectively American, Australian, and combined) was assessed by discrimination via area under the receiver operating characteristic curves (AUC-ROC; this measures the model's ability to distinguish patients who obtained the outcomes from those who did not), calibration metrics (which plot the predicted and the observed probabilities) and Brier score (a composite of discrimination and calibration). In addition, the sensitivity (true positives, recall), specificity (true negatives), positive predictive value (also known as precision), negative predictive value, F1-score (composite of precision and recall), positive likelihood ratio, and negative likelihood ratio were calculated. RESULTS The combined NLP algorithm (the combined Massachusetts and Australian data) achieved excellent performance on independent testing data from Australia (AUC-ROC 0.97 [95% confidence interval 0.87 to 0.99]), Massachusetts (AUC-ROC 0.99 [95% CI 0.80 to 0.99]) and Maryland (AUC-ROC 0.95 [95% CI 0.93 to 0.97]). The NLP developed based on the Massachusetts cohort had excellent performance in the Maryland cohort (AUC-ROC 0.97 [95% CI 0.95 to 0.99]) but worse performance in the Australian cohort (AUC-ROC 0.74 [95% CI 0.70 to 0.77]). CONCLUSION We demonstrated the clinical utility and reproducibility of an NLP algorithm with combined datasets retaining excellent performance in individual countries relative to algorithms developed in the same country alone for detection of incidental durotomy. Further multi-institutional, international collaborations can facilitate the creation of universal NLP algorithms that improve the quality and safety of orthopaedic surgery globally. The combined NLP algorithm has been incorporated into a freely accessible web application that can be found at https://sorg-apps.shinyapps.io/nlp_incidental_durotomy/ . Clinicians and researchers can use the tool to help incorporate the model in evaluating spine registries or quality and safety departments to automate detection of incidental durotomy and optimize prevention efforts. LEVEL OF EVIDENCE Level III, diagnostic study.
Collapse
Affiliation(s)
- Aditya V. Karhade
- Department of Orthopaedic Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| | - Jacobien H. F. Oosterhoff
- Department of Orthopaedic Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
- Department of Orthopaedic Surgery, Amsterdam University Medical Centers, University of Amsterdam, Amsterdam Movement Sciences, the Netherlands
| | - Olivier Q. Groot
- Department of Orthopaedic Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| | - Nicole Agaronnik
- Department of Orthopaedic Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| | - Jeffrey Ehresman
- Department of Neurosurgery, the Johns Hopkins University School of Medicine, Baltimore, MD, USA
| | - Michiel E. R. Bongers
- Department of Orthopaedic Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| | - Ruurd L. Jaarsma
- Department of Orthopaedic and Trauma Surgery, Flinders Medical Centre, Flinders University, Adelaide, Australia
| | - Santosh I. Poonnoose
- Department of Neurosurgery, Flinders Medical Centre, Flinders University, Adelaide, Australia
| | - Daniel M. Sciubba
- Department of Neurosurgery, the Johns Hopkins University School of Medicine, Baltimore, MD, USA
| | - Daniel G. Tobert
- Department of Orthopaedic Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| | - Job N. Doornberg
- Department of Orthopaedic and Trauma Surgery, Flinders Medical Centre, Flinders University, Adelaide, Australia
- Department of Orthopaedic Surgery, University Medical Centre Groningen, University of Groningen, Groningen, the Netherlands
| | - Joseph H. Schwab
- Department of Orthopaedic Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
| |
Collapse
|
16
|
Zhu VJ, Lenert LA, Barth KS, Simpson KN, Li H, Kopscik M, Brady KT. Automatically identifying opioid use disorder in non-cancer patients on chronic opioid therapy. Health Informatics J 2022; 28:14604582221107808. [PMID: 35726687 PMCID: PMC10826411 DOI: 10.1177/14604582221107808] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/15/2022]
Abstract
Background: Using the International Classification of Diseases (ICD) codes alone to record opioid use disorder (OUD) may not completely document OUD in the electronic health record (EHR). We developed and evaluated natural language processing (NLP) approaches to identify OUD from the clinal note. We explored the concordance between ICD-coded and NLP-identified OUD.Methods: We studied EHRs from 13,654 (female: 8223; male: 5431) adult non-cancer patients who received chronic opioid therapy (COT) and had at least one clinical note between 2013 and 2018. Of eligible patients, we randomly selected 10,218 (75%) patients as the training set and the remaining 3436 patients (25%) as the test dataset for NLP approaches.Results: We generated 539 terms representing OUD mentions in clinical notes (e.g., "opioid use disorder," "opioid abuse," "opioid dependence," "opioid overdose") and 73 terms representing OUD medication treatments. By domain expert manual review for the test dataset, our NLP approach yielded high performance: 98.5% for precision, 100% for recall, and 99.2% for F-measure. The concordance of these NLP and ICD identified OUD was modest (Kappa = 0.63).Conclusions: Our NLP approach can accurately identify OUD patients from clinical notes. The combined use of ICD diagnostic code and NLP approach can improve OUD identification.
Collapse
Affiliation(s)
- Vivienne J Zhu
- Biomedical Informatics Center, Department of Public Health Science, College of Medicine, 2345Medical University of South Carolina, Charleston, SC, USA
| | - Leslie A Lenert
- Biomedical Informatics Center, Department of Public Health Science, College of Medicine, 2345Medical University of South Carolina, Charleston, SC, USA
| | - Kelly S Barth
- Department of Psychiatry and Behavioral Science, College of Medicine, 2345Medical University of South Carolina, Charleston, SC, USA
| | - Kit N Simpson
- Department of Healthcare Leadership and Management, College of Health Professions, 2345Medical University of South Carolina, Charleston, SC, USA
| | - Hong Li
- Department of Public Health Science, College of Medicine, 2345Medical University of South Carolina, Charleston, SC, USA
| | - Michael Kopscik
- College of Medicine, 2345Medical University of South Carolina, Charleston, SC, USA
| | - Kathleen T Brady
- Department of Psychiatry and Behavioral Science, College of Medicine, 2345Medical University of South Carolina, Charleston, SC, USA
| |
Collapse
|
17
|
Clinical Text Data Categorization and Feature Extraction Using Medical-Fissure Algorithm and Neg-Seq Algorithm. COMPUTATIONAL INTELLIGENCE AND NEUROSCIENCE 2022; 2022:5759521. [PMID: 35295284 PMCID: PMC8920702 DOI: 10.1155/2022/5759521] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 01/07/2022] [Revised: 01/22/2022] [Accepted: 01/27/2022] [Indexed: 12/19/2022]
Abstract
A large amount of patient information has been gathered in Electronic Health Records (EHRs) concerning their conditions. An EHR, as an unstructured text document, serves to maintain health by identifying, treating, and curing illnesses. In this research, the technical complexities in extracting the clinical text data are removed by using machine learning and natural language processing techniques, in which an unstructured clinical text data with low data quality is recognized by Halve Progression, which uses Medical-Fissure Algorithm which provides better data quality and makes diagnosis easier by using a cross-validation approach. Moreover, to enhance the accuracy in extracting and mapping clinical text data, Clinical Data Progression uses Neg-Seq Algorithm in which the redundancy in clinical text data is removed. Finally, the extracted clinical text data is stored in the cloud with a secret key to enhance security. The proposed technique improves the data quality and provides an efficient data extraction with high accuracy of 99.6%.
Collapse
|
18
|
Carrillo-Larco RM, Castillo-Cara M, Lovón-Melgarejo J. Government plans in the 2016 and 2021 Peruvian presidential elections: A natural language processing analysis of the health chapters. Wellcome Open Res 2022. [DOI: 10.12688/wellcomeopenres.16867.4] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
Background: While clinical medicine has exploded, electronic health records for Natural Language Processing (NLP) analyses, public health, and health policy research have not yet adopted these algorithms. We aimed to dissect the health chapters of the government plans of the 2016 and 2021 Peruvian presidential elections, and to compare different NLP algorithms. Methods: From the government plans (18 in 2016; 19 in 2021) we extracted each sentence from the health chapters. We used five NLP algorithms to extract keywords and phrases from each plan: Term Frequency–Inverse Document Frequency (TF-IDF), Latent Dirichlet Allocation (LDA), TextRank, Keywords Bidirectional Encoder Representations from Transformers (KeyBERT), and Rapid Automatic Keywords Extraction (Rake). Results: In 2016 we analysed 630 sentences, whereas in 2021 there were 1,685 sentences. The TF-IDF algorithm showed that in 2016, 22 terms appeared with a frequency of 0.05 or greater, while in 2021 27 terms met this criterion. The LDA algorithm defined two groups. The first included terms related to things the population would receive (e.g., ’insurance’), while the second included terms about the health system (e.g., ’capacity’). In 2021, most of the government plans belonged to the second group. The TextRank analysis provided keywords showing that ’universal health coverage’ appeared frequently in 2016, while in 2021 keywords about the COVID-19 pandemic were often found. The KeyBERT algorithm provided keywords based on the context of the text. These keywords identified some underlying characteristics of the political party (e.g., political spectrum such as left-wing). The Rake algorithm delivered phrases, in which we found ’universal health coverage’ in 2016 and 2021. Conclusion: The NLP analysis could be used to inform on the underlying priorities in each government plan. NLP analysis could also be included in research of health policies and politics during general elections and provide informative summaries for the general population.
Collapse
|
19
|
Carrillo-Larco RM, Castillo-Cara M, Lovón-Melgarejo J. Government plans in the 2016 and 2021 Peruvian presidential elections: A natural language processing analysis of the health chapters. Wellcome Open Res 2021. [DOI: 10.12688/wellcomeopenres.16867.3] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
Background: While clinical medicine has exploded, electronic health records for Natural Language Processing (NLP) analyses, public health, and health policy research have not yet adopted these algorithms. We aimed to dissect the health chapters of the government plans of the 2016 and 2021 Peruvian presidential elections, and to compare different NLP algorithms. Methods: From the government plans (18 in 2016; 19 in 2021) we extracted each sentence from the health chapters. We used five NLP algorithms to extract keywords and phrases from each plan: Term Frequency–Inverse Document Frequency (TF-IDF), Latent Dirichlet Allocation (LDA), TextRank, Keywords Bidirectional Encoder Representations from Transformers (KeyBERT), and Rapid Automatic Keywords Extraction (Rake). Results: In 2016 we analysed 630 sentences, whereas in 2021 there were 1,685 sentences. The TF-IDF algorithm showed that in 2016, 22 terms appeared with a frequency of 0.05 or greater, while in 2021 27 terms met this criterion. The LDA algorithm defined two groups. The first included terms related to things the population would receive (e.g., ’insurance’), while the second included terms about the health system (e.g., ’capacity’). In 2021, most of the government plans belonged to the second group. The TextRank analysis provided keywords showing that ’universal health coverage’ appeared frequently in 2016, while in 2021 keywords about the COVID-19 pandemic were often found. The KeyBERT algorithm provided keywords based on the context of the text. These keywords identified some underlying characteristics of the political party (e.g., political spectrum such as left-wing). The Rake algorithm delivered phrases, in which we found ’universal health coverage’ in 2016 and 2021. Conclusion: The NLP analysis could be used to inform on the underlying priorities in each government plan. NLP analysis could also be included in research of health policies and politics during general elections and provide informative summaries for the general population.
Collapse
|
20
|
Carrillo-Larco RM, Castillo-Cara M, Lovón-Melgarejo J. Government plans in the 2016 and 2021 Peruvian presidential elections: A natural language processing analysis of the health chapters. Wellcome Open Res 2021. [DOI: 10.12688/wellcomeopenres.16867.2] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
Background: While clinical medicine has exploded, electronic health records for Natural Language Processing (NLP) analyses, public health, and health policy research have not yet adopted these algorithms. We aimed to dissect the health chapters of the government plans of the 2016 and 2021 Peruvian presidential elections, and to compare different NLP algorithms. Methods: From the government plans (18 in 2016; 19 in 2021) we extracted each sentence from the health chapters. We used five NLP algorithms to extract keywords and phrases from each plan: Term Frequency–Inverse Document Frequency (TF-IDF), Latent Dirichlet Allocation (LDA), TextRank, Keywords Bidirectional Encoder Representations from Transformers (KeyBERT), and Rapid Automatic Keywords Extraction (Rake). Results: In 2016 we analysed 630 sentences, whereas in 2021 there were 1,685 sentences. The TF-IDF algorithm showed that in 2016, nine terms appeared with a frequency of 0.10 or greater, while in 2021 43 terms met this criterion. The LDA algorithm defined two groups. The first included terms related to things the population would receive (e.g., ’insurance’), while the second included terms about the health system (e.g., ’capacity’). In 2021, most of the government plans belonged to the second group. The TextRank analysis provided keywords showing that ’universal health coverage’ appeared frequently in 2016, while in 2021 keywords about the COVID-19 pandemic were often found. The KeyBERT algorithm provided keywords based on the context of the text. These keywords identified some underlying characteristics of the political party (e.g., political spectrum such as left-wing). The Rake algorithm delivered phrases, in which we found ’universal health coverage’ in 2016 and 2021. Conclusion: The NLP analysis could be used to inform on the underlying priorities in each government plan. NLP analysis could also be included in research of health policies and politics during general elections and provide informative summaries for the general population.
Collapse
|
21
|
de Oliveira JM, da Costa CA, Antunes RS. Data structuring of electronic health records: a systematic review. HEALTH AND TECHNOLOGY 2021. [DOI: 10.1007/s12553-021-00607-w] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/10/2023]
|
22
|
Olthof AW, Shouche P, Fennema EM, IJpma FFA, Koolstra RHC, Stirler VMA, van Ooijen PMA, Cornelissen LJ. Machine learning based natural language processing of radiology reports in orthopaedic trauma. COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE 2021; 208:106304. [PMID: 34333208 DOI: 10.1016/j.cmpb.2021.106304] [Citation(s) in RCA: 15] [Impact Index Per Article: 5.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 05/22/2020] [Accepted: 07/18/2021] [Indexed: 06/13/2023]
Abstract
OBJECTIVES To compare different Machine Learning (ML) Natural Language Processing (NLP) methods to classify radiology reports in orthopaedic trauma for the presence of injuries. Assessing NLP performance is a prerequisite for downstream tasks and therefore of importance from a clinical perspective (avoiding missed injuries, quality check, insight in diagnostic yield) as well as from a research perspective (identification of patient cohorts, annotation of radiographs). METHODS Datasets of Dutch radiology reports of injured extremities (n = 2469, 33% fractures) and chest radiographs (n = 799, 20% pneumothorax) were collected in two different hospitals and labeled by radiologists and trauma surgeons for the presence or absence of injuries. NLP classification was applied and optimized by testing different preprocessing steps and different classifiers (Rule-based, ML, and Bidirectional Encoder Representations from Transformers (BERT)). Performance was assessed by F1-score, AUC, sensitivity, specificity and accuracy. RESULTS The deep learning based BERT model outperforms all other classification methods which were assessed. The model achieved an F1-score of (95 ± 2)% and accuracy of (96 ± 1)% on a dataset of simple reports (n= 2469), and an F1 of (83 ± 7)% with accuracy (93 ± 2)% on a dataset of complex reports (n= 799). CONCLUSION BERT NLP outperforms traditional ML and rule-base classifiers when applied to Dutch radiology reports in orthopaedic trauma.
Collapse
Affiliation(s)
- A W Olthof
- Department of Radiology, Treant Health Care Group, Dr. G.H. Amshoffweg 1, Hoogeveen, the Netherlands; Department of Radiation Oncology, University Medical Center Groningen, University of Groningen, Hanzeplein 1, Groningen, the Netherlands.
| | - P Shouche
- Department of Radiation Oncology, University Medical Center Groningen, University of Groningen, Hanzeplein 1, Groningen, the Netherlands
| | - E M Fennema
- Department of Trauma Surgery, University Medical Center Groningen, University of Groningen, Hanzeplein 1, Groningen, the Netherlands
| | - F F A IJpma
- Department of Trauma Surgery, University Medical Center Groningen, University of Groningen, Hanzeplein 1, Groningen, the Netherlands
| | - R H C Koolstra
- Department of Radiology, Treant Health Care Group, Dr. G.H. Amshoffweg 1, Hoogeveen, the Netherlands
| | - V M A Stirler
- Department of Trauma Surgery, University Medical Center Groningen, University of Groningen, Hanzeplein 1, Groningen, the Netherlands
| | - P M A van Ooijen
- Department of Radiation Oncology, University Medical Center Groningen, University of Groningen, Hanzeplein 1, Groningen, the Netherlands; Machine Learning Lab, Data Science Center in Health (DASH),University Medical Center Groningen, University of Groningen, L.J. Zielstraweg 2, Groningen, the Netherlands
| | - L J Cornelissen
- Department of Radiation Oncology, University Medical Center Groningen, University of Groningen, Hanzeplein 1, Groningen, the Netherlands; COSMONiO Imaging BV, L.J. Zielstraweg 2, Groningen, the Netherlands
| |
Collapse
|
23
|
Vashishth S, Newman-Griffis D, Joshi R, Dutt R, Rosé CP. Improving broad-coverage medical entity linking with semantic type prediction and large-scale datasets. J Biomed Inform 2021; 121:103880. [PMID: 34390853 PMCID: PMC8952339 DOI: 10.1016/j.jbi.2021.103880] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/30/2021] [Revised: 07/31/2021] [Accepted: 07/31/2021] [Indexed: 10/28/2022]
Abstract
OBJECTIVES Biomedical natural language processing tools are increasingly being applied for broad-coverage information extraction-extracting medical information of all types in a scientific document or a clinical note. In such broad-coverage settings, linking mentions of medical concepts to standardized vocabularies requires choosing the best candidate concepts from large inventories covering dozens of types. This study presents a novel semantic type prediction module for biomedical NLP pipelines and two automatically-constructed, large-scale datasets with broad coverage of semantic types. METHODS We experiment with five off-the-shelf biomedical NLP toolkits on four benchmark datasets for medical information extraction from scientific literature and clinical notes. All toolkits adopt a staged approach of mention detection followed by two stages of medical entity linking: (1) generating a list of candidate concepts, and (2) picking the best concept among them. We introduce a semantic type prediction module to alleviate the problem of overgeneration of candidate concepts by filtering out irrelevant candidate concepts based on the predicted semantic type of a mention. We present MedType, a fully modular semantic type prediction model which we integrate into the existing NLP toolkits. To address the dearth of broad-coverage training data for medical information extraction, we further present WikiMed and PubMedDS, two large-scale datasets for medical entity linking. RESULTS Semantic type filtering improves medical entity linking performance across all toolkits and datasets, often by several percentage points of F-1. Further, pretraining MedType on our novel datasets achieves state-of-the-art performance for semantic type prediction in biomedical text. CONCLUSIONS Semantic type prediction is a key part of building accurate NLP pipelines for broad-coverage information extraction from biomedical text. We make our source code and novel datasets publicly available to foster reproducible research.
Collapse
Affiliation(s)
| | | | - Rishabh Joshi
- Carnegie Mellon University, 5000 Forbes Ave, Pittsburgh, PA, USA
| | - Ritam Dutt
- Carnegie Mellon University, 5000 Forbes Ave, Pittsburgh, PA, USA
| | - Carolyn P Rosé
- Carnegie Mellon University, 5000 Forbes Ave, Pittsburgh, PA, USA
| |
Collapse
|
24
|
[Standardized diagnosis of pancreatic head carcinoma]. DER PATHOLOGE 2021; 42:453-463. [PMID: 34357472 DOI: 10.1007/s00292-021-00971-4] [Citation(s) in RCA: 4] [Impact Index Per Article: 1.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Accepted: 07/07/2021] [Indexed: 10/20/2022]
Abstract
Most pancreatic ductal adenocarcinomas are localized in the pancreatic head. Due to the complex anatomic relationships with the surrounding organs and vascular structures in the retroperitoneal space and to the presence of numerous transection margins and dissection planes, pancreatic head resections belong to the most complex specimens concerning grossing and sampling for histopathologic analysis.Here we discuss current guidelines for standardized grossing and reporting of pancreatic cancer, with special reference to the assessment of the resection margin status. The importance of standardized reporting for the sake of completeness, comprehensibility, comparability, and quality control as well as for the integration of pathology reports in interdisciplinary digital workflows and artificial intelligence applications will be emphasized.
Collapse
|
25
|
Carrillo-Larco RM, Castillo-Cara M, Lovón-Melgarejo J. Government plans in the 2016 and 2021 Peruvian presidential elections: A natural language processing analysis of the health chapters. Wellcome Open Res 2021. [DOI: 10.12688/wellcomeopenres.16867.1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
Background: While clinical medicine has exploded, electronic health records for Natural Language Processing (NLP) analyses, public health, and health policy research have not yet adopted these algorithms. We aimed to dissect the health chapters of the government plans of the 2016 and 2021 Peruvian presidential elections, and to compare different NLP algorithms. Methods: From the government plans (18 in 2016; 19 in 2021) we extracted each sentence from the health chapters. We used five NLP algorithms to extract keywords and phrases from each plan: Term Frequency–Inverse Document Frequency (TF-IDF), Latent Dirichlet Allocation (LDA), TextRank, Keywords Bidirectional Encoder Representations from Transformers (KeyBERT), and Rapid Automatic Keywords Extraction (Rake). Results: In 2016 we analysed 630 sentences, whereas in 2021 there were 1,685 sentences. The TF-IDF algorithm showed that in 2016, nine terms appeared with a frequency of 0.10 or greater, while in 2021 43 terms met this criterion. The LDA algorithm defined two groups. The first included terms related to things the population would receive (e.g., ’insurance’), while the second included terms about the health system (e.g., ’capacity’). In 2021, most of the government plans belonged to the second group. The TextRank analysis provided keywords showing that ’universal health coverage’ appeared frequently in 2016, while in 2021 keywords about the COVID-19 pandemic were often found. The KeyBERT algorithm provided keywords based on the context of the text. These keywords identified some underlying characteristics of the political party (e.g., political spectrum such as left-wing). The Rake algorithm delivered phrases, in which we found ’universal health coverage’ in 2016 and 2021. Conclusion: The NLP analysis could be used to inform on the underlying priorities in each government plan. NLP analysis could also be included in research of health policies and politics during general elections and provide informative summaries for the general population.
Collapse
|
26
|
Carriere J, Shafi H, Brehon K, Pohar Manhas K, Churchill K, Ho C, Tavakoli M. Case Report: Utilizing AI and NLP to Assist with Healthcare and Rehabilitation During the COVID-19 Pandemic. Front Artif Intell 2021; 4:613637. [PMID: 33733232 PMCID: PMC7907599 DOI: 10.3389/frai.2021.613637] [Citation(s) in RCA: 10] [Impact Index Per Article: 3.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/02/2020] [Accepted: 01/08/2021] [Indexed: 01/16/2023] Open
Abstract
The COVID-19 pandemic has profoundly affected healthcare systems and healthcare delivery worldwide. Policy makers are utilizing social distancing and isolation policies to reduce the risk of transmission and spread of COVID-19, while the research, development, and testing of antiviral treatments and vaccines are ongoing. As part of these isolation policies, in-person healthcare delivery has been reduced, or eliminated, to avoid the risk of COVID-19 infection in high-risk and vulnerable populations, particularly those with comorbidities. Clinicians, occupational therapists, and physiotherapists have traditionally relied on in-person diagnosis and treatment of acute and chronic musculoskeletal (MSK) and neurological conditions and illnesses. The assessment and rehabilitation of persons with acute and chronic conditions has, therefore, been particularly impacted during the pandemic. This article presents a perspective on how Artificial Intelligence and Machine Learning (AI/ML) technologies, such as Natural Language Processing (NLP), can be used to assist with assessment and rehabilitation for acute and chronic conditions.
Collapse
Affiliation(s)
- Jay Carriere
- Department of Electrical and Computer Engineering, University of Alberta, Edmonton, AB, Canada
| | - Hareem Shafi
- Department of Electrical and Computer Engineering, University of Alberta, Edmonton, AB, Canada
| | - Katelyn Brehon
- School of Public Health, University of Alberta, Edmonton, AB, Canada
| | - Kiran Pohar Manhas
- Neurosciences, Rehabilitation, and Vision Strategic Clinical Network, Alberta Health Services, Calgary, AB, Canada
| | - Katie Churchill
- Department of Occupational Therapy, University of Alberta, Edmonton, AB, Canada.,Cumming School of Medicine, University of Calgary, Calgary, AB, Canada
| | - Chester Ho
- Neurosciences, Rehabilitation, and Vision Strategic Clinical Network, Alberta Health Services, Calgary, AB, Canada.,Faculty of Medicine and Dentistry, University of Alberta, Edmonton, AB, Canada
| | - Mahdi Tavakoli
- Department of Electrical and Computer Engineering, University of Alberta, Edmonton, AB, Canada
| |
Collapse
|