A5003609908- Biomedical Text Mining and Ontologies
- Topic Modeling
- Natural Language Processing Techniques
- Electronic Health Records Systems
- Machine Learning in Healthcare
- Pharmaceutical Practices and Patient Outcomes
- Mental Health via Writing
- Blood Pressure and Hypertension Studies
- Patient-Provider Communication in Healthcare
- Nursing Diagnosis and Documentation
- Data Quality and Management
- Semantic Web and Ontologies
- AI-based Problem Solving and Planning
- Patient Satisfaction in Healthcare
- Clinical Reasoning and Diagnostic Skills
- Dementia and Cognitive Impairment Research
- Authorship Attribution and Profiling
- Geriatric Care and Nursing Homes
- Mental Health and Patient Involvement
- Primary Care and Health Outcomes
- Advanced Text Analysis Techniques
University of Manchester
2012-2020
The Christie NHS Foundation Trust
2015-2017
Identification of clinical events (eg, problems, tests, treatments) and associated temporal expressions dates times) are key tasks in extracting managing data from electronic health records. As part the i2b2 2012 Natural Language Processing for Clinical Data challenge, we developed evaluated a system to automatically extract narratives. The extracted were additionally normalized by assigning type, value, modifier.The combines rule-based machine learning approaches that rely on morphological,...
A recent promise to access unstructured clinical data from electronic health records on large-scale has revitalized the interest in automated de-identification of notes, which includes identification mentions Protected Health Information (PHI). We describe methods developed and evaluated as part i2b2/UTHealth 2014 challenge identify PHI defined by 25 entity types longitudinal narratives. Our approach combines knowledge-driven (dictionaries rules) data-driven (machine learning) with a large...
Abstract Objectives As electronic mental health records become more widely available, several approaches have been suggested to automatically extract information from free‐text narrative aiming support epidemiological research and clinical decision‐making. In this paper, we explore extraction of explicit mentions symptom severity initial psychiatric evaluation records. We use the data provided by 2016 CEGS N‐GRID NLP shared task Track 2, which contains 541 manually annotated for according...
Heart disease is the leading cause of death globally and a significant part human population lives with it. A number risk factors have been recognized as contributing to disease, including obesity, coronary artery (CAD), hypertension, hyperlipidemia, diabetes, smoking, family history premature CAD. This paper describes evaluates methodology extract mentions such from diabetic clinical notes, which was task i2b2/UTHealth 2014 Challenge in Natural Language Processing for Clinical Data. The...
Background The efficacy of acetylcholinesterase inhibitors and memantine in the symptomatic treatment Alzheimer's disease is well-established. Randomised trials have shown them to be associated with a reduction rate cognitive decline. Aims To investigate real-world effectiveness for dementia-causing diseases largest UK observational secondary care service data-set date. Method We extracted mentions relevant medications testing (Mini-Mental State Examination (MMSE) Montreal Cognitive...
The way we collect and use patient experience data is vital to optimise the quality safety of health services. Yet, some patients carers do not give feedback because limited ways collected, analysed presented. In this study, worked together with researchers, staff, carer participants, public involvement engagement (PPIE) contributors, co-design new tools for collection in multiple settings. This paper outlines how range PPIE research activities enabled data.Eight contributors represented a...
In addition to structured data, electronic health records contain unstructured clinical notes and narratives. The identification classification of mentions relevant concepts is a crucial preprocessing step in designing developing decision support systems. While this task has gained significant attention recent years, there are still number issues that need further investigation. This paper explores variety common challenges faced by named entity recognition methods as well current approaches...
We describe and evaluate an automated approach used as part of the i2b2 2011 challenge to identify categorise statements in suicide notes into one 15 topics, including Love, Guilt, Thankfulness, Hopelessness Instructions. The combines a set lexico-syntactic rules with models derived by machine learning from training dataset. rely on named entities, lexical, lexico-semantic presentation features, well that are applicable given statement. On testing 300 notes, showed overall best micro...
Background Collecting NHS patient experience data is critical to ensure the delivery of high-quality services. Data are obtained from multiple sources, including service-specific surveys and widely used generic surveys. There concerns about timeliness feedback, that some groups patients carers do not give feedback free-text may be useful but difficult analyse. Objective To understand how improve collection usefulness in services for people with long-term conditions using digital capture...
The problem of named entity recognition in the medical/clinical domain has gained increasing attention do to its vital role a wide range clinical decision support applications. identification complete and correct term span is for further knowledge synthesis (e.g., coding/mapping concepts thesauruses classification standards). This paper investigates boundary adjustment by sequence labeling representations models post-processing techniques (recognition events). Using current state-of-the-art...
This report describes a minimalistic set of methods engineered to anchor clinical events onto temporal space. Specifically, we describe extract (e.g., Problems, Treatments and Tests), expressions (i.e., time, date, duration, frequency), links Before, After, Overlap) between entities. These are developed validated using high quality datasets.
ABSTRACTObjectivesIncreasing interest to use unstructured electronic patient records for research has attracted attention automated de-identification methods conduct large scale removal of Personal Identifiable Information (PII). PII mainly include identifiable information such as person names, dates (e.g., date birth), reference numbers hospital number, NHS number), locations addresses), contacts telephone, e-mail), occupation, age, and other identity (ethnical, religion, sexual) mentioned...