This research project explored an innovative method to retrieve clinically relevant images for facilitating timely and accurate evaluation of diabetic retinopathy.
Developing Evidence-Based, User-Centered Design and Implementation Guidelines to Improve Health Information Technology Usability
Analysis of 5,200 patient safety event reports showed an association between electronic health record (EHR) usability and patient safety in both adults and children and led to development of an EHR usability and safety assessment tool that healthcare facilities can use to identify usability and safety issues.
This project developed and pilot-tested a novel, outcomes-based emergency department triage tool and found that risk stratification and waiting times were improved for some patients.
This research studied errors in medical documents created with speech recognition software and developed natural language processing methods to detect such errors.
This research evaluated the implementation and effectiveness of a clinical decision support tool designed to support the delivery of recommended care to hospitalized patients with heart failure, regardless of the reason for hospitalization.
This research took an existing sepsis-related clinical decision support (CDS) and developed, tested, implemented, and validated a knowledge-based artificial intelligence-enhanced sepsis CDS.
Using the Electronic Health Record To Identify Children Likely To Suffer Last-Minute Surgery Cancellation
This research applied machine learning to develop a model predicting surgical cancellations among pediatric patients, and found the feasibility in using these algorithms as a cost-effective quality-improvement measure.
Natural Language Processing To Identify and Rank Clinically Relevant Information for EHRs in the Emergency Department
This project developed a natural language processing electronic health record search tool that automatically identifies and ranks relevant clinical information based on a patient’s presenting complaint within the emergency department setting.
This research prospectively evaluated a machine learning algorithm that identifies candidates for neurologic surgery to control epilepsy.