A5091282584- Machine Learning in Healthcare
- Biomedical Text Mining and Ontologies
- Mental Health Research Topics
- Mental Health via Writing
- Suicide and Self-Harm Studies
- Sepsis Diagnosis and Treatment
- Topic Modeling
- Treatment of Major Depression
- Food Security and Health in Diverse Populations
- Health disparities and outcomes
- Chronic Kidney Disease and Diabetes
- Mental Health Treatment and Access
- Zeolite Catalysis and Synthesis
- Electronic Health Records Systems
- Nursing Diagnosis and Documentation
- Acute Kidney Injury Research
- Clinical practice guidelines implementation
- Chronic Disease Management Strategies
- Digital Mental Health Interventions
- Dementia and Cognitive Impairment Research
- Migration, Health and Trauma
- Maternal Mental Health During Pregnancy and Postpartum
- Attachment and Relationship Dynamics
- Catalysis and Oxidation Reactions
- Global Cancer Incidence and Screening
Cornell University
2018-2025
Weill Cornell Medicine
2018-2024
Institute of Information Technologies
2021
Northwestern University
2019
Auburn University
2003-2004
We aimed to address deficiencies in structured electronic health record (EHR) data for race and ethnicity by identifying black Hispanic patients from unstructured clinical notes assessing differences between with or without race/ethnicity data.Using EHR 16 665 encounters at a primary care practice, we developed rule-based natural language processing (NLP) algorithms classify as black/Hispanic. evaluated performance of the method against an annotated gold standard, compared NLP-derived data,...
Abstract Objectives Social support (SS) and social isolation (SI) are determinants of health (SDOH) associated with psychiatric outcomes. In electronic records (EHRs), individual-level SS/SI is typically documented in narrative clinical notes rather than as structured coded data. Natural language processing (NLP) algorithms can automate the otherwise labor-intensive process extraction such information. Materials Methods Psychiatric encounter from Mount Sinai Health System (MSHS, n = 300)...
Abstract Introduction We sought to assess longitudinal electronic health records (EHRs) using machine learning (ML) methods computationally derive probable Alzheimer's Disease (AD) and related dementia subphenotypes. Methods A retrospective analysis of EHR data from a cohort 7587 patients seen at large, multi‐specialty urban academic medical center in New York was conducted. Subphenotypes were derived hierarchical clustering 792 AD (cases) who had received least one diagnosis their clinical...
Left ventricular ejection fraction (EF) is a key measure in the diagnosis and treatment of heart failure (HF) many patients experience changes EF overtime. Large-scale analysis longitudinal using electronic health records (EHRs) limited. In multi-site retrospective study EHR data from three academic medical centers, we investigated measurements diagnosed with HF. We observed significant variations baseline characteristics change behavior HF cohorts previous that based on registry data. Data...
Social and behavioral determinants of health (SBDoH) have important roles in shaping people's health. In clinical research studies, especially comparative effectiveness failure to adjust for SBDoH factors will potentially cause confounding issues misclassification errors either statistical analyses machine learning-based models. However, there are limited studies examine outcomes due the lack structured information current electronic record (EHR) systems, while much is documented narratives....
Objective To evaluate if a machine learning approach can accurately predict antidepressant treatment outcome using electronic health records (EHRs) from patients with depression. Method This study examined 808 depression at New York City‐based outpatient mental clinic between June 13, 2016 and 22, 2020. Antidepressant was defined based on trend in symptom severity over time categorized as either "Recovering" or "Worsening" (i.e., non‐Recovering), measured by the slope of individual‐level...
Background: Assessing treatment response in patients with myeloproliferative neoplasms is difficult because data components exist unstructured bone marrow pathology (hematopathology) reports, which require specialized, manual annotation and interpretation. Although natural language processing (NLP) has been successfully implemented for the extraction of features from solid tumor little known about its application to hematopathology. Methods: An open-source NLP framework called Leo was parse...
Abstract Introduction Electronic health record (EHR)‐driven phenotyping is a critical first step in generating biomedical knowledge from EHR data. Despite recent progress, current approaches are manual, time‐consuming, error‐prone, and platform‐specific. This results duplication of effort highly variable across systems institutions, not scalable or portable. In this work, we investigate how the nascent Clinical Quality Language (CQL) can address these issues enable high‐throughput,...
Abstract Objective To identify depression subphenotypes from Electronic Health Records (EHRs) using machine learning methods, and analyze their characteristics with respect to patient demographics, comorbidities, medications. Materials Methods Using EHRs the INSIGHT Clinical Research Network (CRN) database, multiple (ML) algorithms were applied 11 275 patients discern distinct characteristics. Results computational approaches, we derived three subphenotypes: Phenotype_A (n = 2791; 31.35%)...
Abstract The objective of this study was to investigate the potential association between use four frequently prescribed drug classes, namely antihypertensive drugs, statins, selective serotonin reuptake inhibitors, and proton-pump likelihood disease progression from mild cognitive impairment (MCI) dementia using electronic health records (EHRs). We conducted a retrospective cohort observational EHRs approximately 2 million patients seen at large, multi-specialty urban academic medical...
Acute Kidney Injury (AKI) is the most common cause of organ dysfunction in critically ill adults and prior studies have shown AKI associated with a significant increase mortality risk. Early prediction risk for patients can help clinical decision makers better understand patient condition time take appropriate actions. However, heterogeneous disease its complex, which makes such predictions challenging task. In this paper, we investigate machine learning models predicting who are stratified...