Search AMIA.org

This study aimed to develop a novel multi-stage self-supervised learning model tailored for the accurate classification of optical coherence tomography (OCT) images in ophthalmology reducing reliance on costly labeled datasets while maintaining high diagnostic accuracy.

(1) Describe the evolution of Health Information Exchanges (HIEs) into Health Data Utilities (HDUs); (2) Provide motivation for HDUs as a public strategic investment target.

Although efforts to effectively govern AI continue to develop, relatively little work has been done to systematically measure and include patient perspectives or expectations of AI in governance. This analysis is designed to understand patient expectations of healthcare AI.

Direct electronic access to multiple electronic health record (EHR) systems through patient portals offers a novel avenue for decentralized research. Given the critical value of patient characterization, we sought to compare computable evaluation of health conditions from patient-portal EHR against the traditional self-report.

This study aimed to assess the efficacy of a biomedical informatics boot camp with regard to improving the skill sets of its participants.The University at Buffalo hosts a free, virtual biomedical informatics boot camp annually. Lectures covering various subject matters are offered, for example, general programming, machine learning, natural language processing, and clinical decision support. Once the 2023 boot camp had concluded, an anonymous voluntary survey was distributed.Seventy percent of [...]

Despite the proven usefulness of appropriate clinical decision support (CDS) alerts, many CDS systems fire excessive, clinically irrelevant alerts that are often ignored by clinicians. We have developed a method to suppress false-positive alerts based on prior drug tolerance but encountered substantial barriers to integrating the method into widely adopted commercial electronic health record (EHR) systems.This study aimed to describe the challenges faced while attempting to integrate our method into [...]

At a large quaternary health system, tissue specimens were frequently sent to the microbiology laboratory with an incorrect wound culture order meant for swab specimens due to poor electronic health record (EHR) menu design. Wound cultures were also requested in chronic wound cases with a low index of suspicion for acute infection.This study aimed to present a case report on specific changes to the design of the electronic test menu [...]

This study assesses the abilities of 2 large language models (LLMs), GPT-4 and BioMistral 7B, in responding to patient queries, particularly concerning rare diseases, and compares their performance with that of physicians.

Pediatric cancer patients with fever and neutropenia are at risk for bacterial sepsis, traditionally requiring extended hospital stays on antibiotics until neutrophil counts recover. According to a newly validated scoring system, a subset of these patients is at lower risk and eligible for early discharge and reduced intravenous (IV) antibiotic exposure.Reduce length-of-stay (LOS) for febrile neutropenic patients using clinical decision support (CDS) to identify low-risk patients.A CDS system was developed [...]