A5058000194- Chronic Kidney Disease and Diabetes
- Biomedical Text Mining and Ontologies
- Advanced Causal Inference Techniques
- Healthcare Policy and Management
- Health Systems, Economic Evaluations, Quality of Life
- Liver Disease Diagnosis and Treatment
- Machine Learning and Data Classification
- Statistical Methods in Clinical Trials
- Machine Learning in Healthcare
- Chronic Disease Management Strategies
- Semantic Web and Ontologies
- Electronic Health Records Systems
- Diet and metabolism studies
- Imbalanced Data Classification Techniques
- Advanced Data Storage Technologies
- Colorectal Cancer Screening and Detection
- Bariatric Surgery and Outcomes
- MicroRNA in disease regulation
- Migration, Health and Trauma
- Liver physiology and pathology
- Global Cancer Incidence and Screening
- Interpreting and Communication in Healthcare
- Dialysis and Renal Disease Management
- Scientific Computing and Data Management
Columbia University Irving Medical Center
2021-2024
Columbia University
2020-2024
As the prevalence of obesity-induced type 2 diabetes mellitus (T2DM) and nonalcoholic steatohepatitis (NASH) continue to increase, need for pharmacologic therapies becomes urgent. However, endeavors identify develop novel therapeutic strategies these chronic conditions are balanced by safety, impeding clinical translation. One shared pathology two diseases is a maladaptive reactivation Notch signaling pathway in liver. antagonism with γ-secretase inhibitors effectively suppresses hepatic...
Abstract Objectives Chart review as the current gold standard for phenotype evaluation cannot support observational research on electronic health records and claims data sources at scale. We aimed to evaluate ability of structured efficient interpretable an alternative chart review. Materials Methods developed Knowledge-Enhanced Electronic Profile Review (KEEPER) a tool that extracts patient’s elements relevant presents them in standardized fashion following clinical reasoning principles....
Healthcare continues to grapple with the persistent issue of treatment disparities, sparking concerns regarding equitable allocation treatments in clinical practice. While various fairness metrics have emerged assess decision-making processes, a growing focus has been on causality-based concepts due their capacity mitigate confounding effects and reason about bias. However, application causal notions evaluating electronic health record (EHR) data remains an understudied domain. This study...
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 With the burgeoning development of computational phenotypes, it is increasingly difficult to identify right phenotype for tasks. This study uses a mixed-methods approach develop and evaluate novel metadata framework retrieval reusing phenotypes. Twenty active phenotyping researchers from 2 large research networks, Electronic Medical Records Genomics Observational Health Data Sciences Informatics, were recruited suggest elements. Once consensus was reached on 39 elements, 47 new...
Data-driven clinical prediction algorithms are used widely by clinicians. Understanding what factors can impact the performance and fairness of data-driven is an important step towards achieving equitable healthcare. To investigate modeling choices on algorithmic fairness, we make use a case study to build algorithm for estimating glomerular filtration rate (GFR) based patient's electronic health record (EHR). We compare three distinct approaches GFR: CKD-EPI equations, epidemiological...
Data-driven clinical prediction algorithms are used widely by clinicians. Understanding what factors can impact the performance and fairness of data-driven is an important step towards achieving equitable healthcare. To investigate modeling choices on algorithmic fairness, we make use a case study to build algorithm for estimating glomerular filtration rate (GFR) based patient's electronic health record (EHR). We compare three distinct approaches GFR: CKD-EPI equations, epidemiological...
Type 2 diabetes mellitus is a complex and under-treated disorder closely intertwined with obesity. Adolescents severe obesity type have more aggressive disease compared to adults, rapid decline in pancreatic β cell function increased incidence of comorbidities. Given the relative paucity pharmacotherapies, bariatric surgery has become increasingly used as therapeutic option. However, subsets this population sub-optimal outcomes either inadequate weight loss or little improvement disease....
Healthcare continues to grapple with the persistent issue of treatment disparities, sparking concerns regarding equitable allocation treatments in clinical practice. While various fairness metrics have emerged assess decision-making processes, a growing focus has been on causality-based concepts due their capacity mitigate confounding effects and reason about bias. However, application causal notions evaluating electronic health record (EHR) data remains an understudied domain. This study...
Abstract Background The appropriate use and the implications of using variables that attempt to encode a patient’s race in clinical predictive algorithms remain unclear. algorithm for estimating glomerular filtration rate (GFR) adjusts race, but observed difference between Black non-Black participants lacks biologically substantiated evidence. We investigated impact variable on GFR prediction by race-stratified error analysis. Methods implemented three with varied amount input information...