Innovations for Clinical & Translational Impact

GENiUS is an AI-powered platform designed to end long diagnostic odysseys for the most vulnerable patients. Built on AWS and seamlessly integrated into the Epic electronic health record (EHR) system, the platform identifies which infants in the neonatal intensive care unit (NICU) need rapid genome sequencing, automates the complex interpretation of genomic data and continuously reanalyzes past cases as new scientific discoveries emerge about the underlying genetic causes of complex medical conditions.

Currently, the project focuses on rapid genetic testing and diagnosis among the highest-acuity patients in the NICU using the NeoGx platform built by ODS. This platform can extract detailed clinical information from patients’ EHRs to compute phenotypes using AI-guided natural language processing based on patients’ signs and symptoms, test results and developmental progression.

This work extends ODS’s contributions to AI-enabled clinical genomics at Nationwide Children’s that are directly improving patient outcomes and advancing precision medicine. These contributions include the development of a state-of-the-art, cloud-based genomic data warehouse equipped with a powerful interface that allows users to filter, prioritize, annotate, interpret and automatically rescreen variants identified through rapid genome sequencing.

Goals and Outcomes:

• Cut diagnostic delays by more than 50% and reduce time to the first genetic test by 90%.
• Increase the number of definitive genetic diagnoses made within two weeks of testing by a factor of two to four.
• Stage the platform for deployment in the Level IV NICU, with active development efforts to expand across NICU and primary care settings.
• Earn competitive funding awards from the National Institutes of Health and AWS.

Responding to the escalating pediatric mental health crisis, the DREAM initiative deploys cutting-edge generative AI to empower researchers at the nation’s largest Pediatric Behavioral Health program. This initiative, in collaboration with Eric Youngstrom, PhD, focuses on unlocking the potential of the Longitudinal Assessment of Manic Symptoms (LAMS) dataset to provide deeper insights into pediatric bipolar spectrum disorders.

Researchers have historically struggled to navigate vast, fragmented datasets to uncover root causes and identify timely, effective interventions. The DREAM project addresses this by providing an AI-powered tool that democratizes access to these complex datasets, allowing researchers to ask sophisticated questions and extract insights without needing deep technical expertise. This accelerates the identification of effective interventions for youth mental health and underscores the ODS’s commitment to improving analytical capabilities to solve the most pressing issues in pediatric care. 

• AI-Powered Data Analysis: An AI-powered chatbot that helps researchers efficiently extract insights from complex mental health datasets.
• Natural Language Querying: Allows researchers to query datasets in natural languages, bypassing the need for extensive database knowledge and pre-processing.
• Advanced Data Retrieval:Employs a methodology that leverages parallelized interactions with LLMs to accurately surface relevant data, even with semantic variations
• Streamlined Workflows: Generates free-form code, responds to analytical queries and extract customized data slices to enable more efficient research.

Goals and Outcomes:

• Recognized with prestigious Gartner Eye on Innovation Award for Mid-Size Enterprises (Americas region), highlighting ODS’s leadership in driving digital transformations that address critical research and health care challenges
• Vision includes extending to other enterprise data sources, including electronic health records and genomic sequencing, via integration into the Data Lake environment

The Heart Center Precision Health Program at Nationwide Children's is building one of the first governed, integrated data platforms in pediatric cardiovascular medicine — the Heart Center Data Lake — to transform how complex congenital and acquired heart disease is understood and treated. Serving as the analytic engine of the program, the Heart Center Data Lake will unify currently siloed data streams — genomic sequencing, electronic health records, cardiac imaging and physiologic monitoring, surgical and procedural information, national registry records and longitudinal outcomes — into a scalable, reusable resource designed for discovery, predictive modeling, quality improvement and clinical decision support. Built within a secure cloud environment, the platform will provide investigators with access to multimodal datasets, advanced analytics and AI models curated through an agentic user interface. It is a flagship use case within ODS’s Data Lake and Data Translation strategic pillars.

The Data Lake will function as the R&D backbone of a broader bench-to-bedside innovation ecosystem that targets the most consequential and least-understood conditions in pediatric cardiology. This effort is anchored by COURAGE for Kids, Nationwide Children's prospective genome-first cardiovascular study that has already achieved a 35% diagnostic yield across more than 370 enrolled participants since 2023. By linking genomic findings to clinical trajectories and outcomes, the Data Lake will enable a shift beyond diagnosis toward genomically informed outcome prediction. Long-term, the Precision Health Program will foster the creation and implementation of clinical decision support tools, state-of-the-art disease models and interactive dashboards for real-time intelligence at the point of care. Together, these will enable clinicians to intervene earlier, personalize therapies better and anticipate complications faster, achieving best outcomes for children with complex cardiovascular disease and establishing Nationwide Children's as a national destination for data-driven pediatric care.

ODS is spearheading efforts to improve outcomes for critically ill children using AI paires with the GOSSIS, a new open-source algorithm created by the GOSSIS Consortium to assess severity of illness and predict in-hospital mortality in intenisve care unit (ICU) patients. The GOSSIS incorporates diverse predictor variables from patient EHRs, including demographics, medical history, and measures related to their ICU hospital course (diagnoses, vitals, labs, functional status). By extending this clinically validated method to the NICU at Nationwide Children's, ODS has generated a model that predicts mortality on par with its performance in adult and non-neonatial pediatric populations outside Nationwide Children's. This work, carried out in partnership with researchers at the University of Pittsburgh Medical Center, is advancing the cutting edge of critical care across health systems, leveraging AI to help reduce mortality and improve resource allocation among the sickest patients while promoting open-source AI development, data interoperability and multi-institutional collaboration.