A5012523508- Machine Learning in Healthcare
- Topic Modeling
- Biomedical Text Mining and Ontologies
- Artificial Intelligence in Healthcare and Education
- Sex and Gender in Healthcare
- Global Cancer Incidence and Screening
- Diversity and Career in Medicine
- Food Security and Health in Diverse Populations
- Chronic Disease Management Strategies
- Antibiotic Resistance in Bacteria
- Migration, Health and Trauma
- Antibiotics Pharmacokinetics and Efficacy
- LGBTQ Health, Identity, and Policy
- Advanced Causal Inference Techniques
- Mobile Health and mHealth Applications
- Healthcare Policy and Management
- Clinical Reasoning and Diagnostic Skills
- Biomedical and Engineering Education
- Bacterial Identification and Susceptibility Testing
- Ethics in Clinical Research
- Interpreting and Communication in Healthcare
- Health Policy Implementation Science
University of Washington Medical Center
2023-2025
University of Washington
2023-2024
Columbia University
2018-2022
Stanford University
2018
Abstract Background Social and behavioral determinants of health (SBDH) are environmental factors that often impede disease management result in sexually transmitted infections. Despite their importance, SBDH inconsistently documented electronic records (EHRs) typically collected only an unstructured format. Evidence suggests structured data elements present EHRs can contribute further to identify the patient record. Objective Explore automated inference both presence documentation...
Polymyxins are last-resort antibiotics used to treat highly resistant Gram-negative bacteria. There increasing reports of polymyxin resistance emerging, raising concerns a postantibiotic era. Polymyxin is therefore significant public health threat, but current phenotypic methods for detection difficult and time-consuming perform. have been efforts use whole-genome sequencing antibiotic resistance, this has apply because its complex polygenic nature. The significance our research that we...
Unstructured clinical narratives are continuously being recorded as part of delivery care in electronic health records, and dedicated tagging staff spend considerable effort manually assigning codes for billing purposes. Despite these efforts, however, label availability accuracy both suboptimal. In this retrospective study, we aimed to automate the assignment top-level International Classification Diseases version 9 (ICD-9) records from human veterinary data stores using minimal manual...
Developing equitable, sustainable informatics solutions is key to scalability and long-term success for projects in the global health (GHI) domain. This paper presents strategies incorporating principles of equity GHI project lifecycle. The American Medical Informatics Association (AMIA) Working Group organized a collaborative workshop at 2023 AMIA Annual Symposium that included presentation five case studies how have been incorporated into situated low-and-middle-income countries with...
Abstract Background Integrated clinical databases from national biobanks have advanced the capacity for disease research. Data quality and completeness filters are used when building cohorts to address limitations of data missingness. However, these may unintentionally introduce systemic biases they correlated with race ethnicity. Objective In this study, we examined ethnicity introduced by applying common 4 records databases. Specifically, evaluated whether that disproportionately exclude...
Many approaches in biomedical informatics (BMI) rely on the ability to define, gather, and manipulate data support health through a cyclical research-practice lifecycle. Researchers within this field are often fortunate work closely with healthcare public systems influence generation capture have access vast amount of data. informaticists also expertise engage stakeholders, develop new methods applications, policy. However, research policy that explicitly seeks address systemic drivers would...
Clinical notes present a wealth of information for applications in the clinical domain, but heterogeneity across institutions and settings presents challenges their processing. The natural language processing field has made strides overcoming domain heterogeneity, while pretrained deep learning models opportunities to transfer knowledge from one task another. Pretrained have performed well when transferred new tasks; however, it is not understood if these generalize differences within...
Observational health research often relies on accurate and complete race ethnicity (RE) patient information, such as characterizing cohorts, assessing quality/performance metrics of hospitals systems, identifying disparities. While the electronic record contains structured data accessible patient-level RE data, it is missing, inaccurate, or lacking granular details. Natural language processing models can be trained to identify in clinical text which supplement missing repositories. Here we...
ABSTRACT Objective Integrated clinical databases from national biobanks have advanced the capacity for disease research. Data quality and completeness filters are used when building cohorts to address limitations of data missingness. However, these may unintentionally introduce systemic biases they correlated with race ethnicity. In this study, we examined race/ethnicity introduced by applying common four records databases. Materials Methods We 19 commonly in electronic health research on...
Abstract Objective The American Medical Informatics Association (AMIA) Task Force on Diversity, Equity, and Inclusion (DEI) was established to address systemic racism health disparities in biomedical informatics, aligning with AMIA’s mission transform healthcare. DEI initiatives were spurred by member voices responding police brutality COVID-19’s impact Black/African communities. Materials Methods Force, consisting of 20 members across 3 groups aligned 2020-2025 Strategic Plan, met biweekly...
ABSTRACT Objective Currently, dedicated tagging staff spend considerable effort assigning clinical codes to patient summaries for public health purposes, and machine-learning automated is bottlenecked by availability of electronic medical records. Veterinary records, a largely untapped data source that could benefit both human non-human patients, fill the gap. Materials Methods In this retrospective study, we trained long short-term memory (LSTM) recurrent neural networks (RNNs) on 52,722...
Background: In public health research, there is currently a need to close the gap between care delivery and cohort identification. We dedicated tagging staff allocate considerable amount of effort assigning clinical codes after reading patient summaries. Machine learning automation can facilitate classification these narratives, but sufficient availability electronic medical records still bottleneck. Veterinary represent largely untapped data source that could be used benefit both human...