Applied Clinical Informatics

National Library Of Medicine

Investigators

Linked publications, trials & patents

Abstract

The Lister Hill National Center for Biomedical Computing (LHNCBC) Applied Clinical Informatics Branch (ACIB) achieves its goals through the following teams, which achieved the following in this reporting period: Clinical Data Management and Analytics --------------------------------------------------- During the past year, relevant work has included ongoing enhancements to the Center for Clinical Observational Investigations (CCOI), which was redesigned to feature a “What’s New” highlight reel informing visitors of new dataset profiles and webinar information as well as the ability to subscribe to the CCOI listerv. An events webpage was added that lists workshop and webinar educational opportunities for those interested in learning about the datasets profiled in the CCOI website and their use. Seven new dataset profiles have been added to CCOI in FY25, for a total of 11 datasets currently profiled on the website. A needs assessment and focus groups have been conducted to ascertain biomedical researcher needs when using clinical data in research and training sessions were held on two of the profiled datasets: the Department of Transportation’s the National Emergency Medical Services Information System (NEMSIS) and the National Institute on Aging’s AgingResearchBiobank. In addition, ACIB has continued to advance the use of health data standards through the development and maintenance of relevant tools. ACIB continues to provide support for RxNorm tools in support of the adoption of RxNorm, NLM's standard terminology for drugs. Enhancements made to RxNorm tools include enabling third-party application use, updates with the aid of new automation for increased coverage and consistency, streamlined search feature to facilitate people's most common searches with fewer gestures, enhanced support for classifications with semantic codes supporting most effective. Part of this work will be presented at the SNOMED Expo 2025. The ACIB team also develops and maintains tools to support use of Systematized Nomenclature of Medicine—Clinical Terms (SNOMED-CT®), the standard nomenclature of Medicine. Ongoing work supports the creation and maintenance of mappings between SNOMED CT and ICD-10-CM to facilitate interoperability between clinical and administrative coding systems. A new project called Intelligent UMLS Editing Assistance (IUEA) was launched to leverage machine learning and AI techniques to identify synonymous medical terms to be tested in the UMLS editing environment. Clinical Data Interoperability ------------------------------------- ACIB continued to participate in the development of the Fast Healthcare Interoperability Resources (FHIR®) standard, though workgroups and through enhancing its suite of tools in support of use of FHIR, and a FHIR Team Lead was hired to oversee these efforts. The LHC-Forms tool, which is used to render and fill out healthcare questionnaires, now supports new features in various versions of FHIR standards. Notable additions include enhanced display options, support for multiple languages, and improved validation and error messaging to ensure the accuracy and usability of the forms. The NLM Form Builder, a tool for designing healthcare questionnaires, has been enhanced with a new editor for creating complex “expressions” (used for calculated form field values) and support for additional data structures. It also includes new features for better data entry, initial values, and improved error handling and accessibility. The Research Data Finder (RDF) tool, used to search healthcare data servers, now includes a new configuration file for customization, support for additional data types, and better filtering capabilities for medication-related resources. Updates to the fhirpath.js JavaScript package, which processes FHIRPath (an “expression” language for calculating values), have introduced new functions and support for additional data types. Additionally, a new cross-team project, “NLM Scrubber on FHIR” was launched to build a system that will provide researchers with access to deidentified data from FHIR-enabled Electronic Healthcare Records (EHR) systems. Development of Technology Solutions ------------------------------------------------ ACIB staff continued to advance processing of clinical images across a variety of activities and areas of applications. This included continuing the development of deep learning algorithms to classify various eye images (e.g., fundus images) across several eye conditions including uveitis, glaucoma, and macular degeneration. Part of this work has resulted in the National Stage filing of the intellectual property models, entitled “SYSTEMS AND METHODS FOR DEEP LEARNING TO DETECT RETINAL VASCULITIS ON COLOR FUNDUS PHOTOGRAPHS OF PATIENTS WITH UVEITIS,” to the United States Patent and Trademark Office. Relatedly, work was conducted to develop prototypes of mobile applications for Android and iOS to capture images and videos, detect optic disc within them in real time, crop the optic disc area, and estimate the probability of glaucoma. In addition, ACIB is advancing the development of deep learning models and software to facilitate screening of infections and infectious diseases such as malaria as well as complications for patients diagnosed with HIV and tuberculosis.