Min Gu Kwak, machine learning research scientist at the PittNAIL (Clinical Natural Language Processing and Artificial Intelligence Innovation Laboratory) lab, has been awarded a fellowship in the AIM-AHEAD (Artificial Intelligence/Machine Learning Consortium to Advance Health Equity and Researcher Diversity) Fast Healthcare Interoperability Resources (FHIR) Collaborative Training Program. Funded by the National Institutes of Health, the training program award is designed to train, accelerate and sustain a multidisciplinary workforce proficient in FHIR and related health data technologies across academia, industry, health care and community-based organizations.
The competitive fellowship runs for 19 weeks and includes a $4,000 stipend. It also marks the second AIM-AHEAD program award for PittNAIL, demonstrating the lab’s continued excellence and recognition on the national stage in health informatics and AI research. The award also reinforces the University of Pittsburgh’s leadership in combining AI with clinical data infrastructure, positioning the University as a hub for innovative health care research.
“I am thrilled to see another PittNAIL lab member receive an AIM-AHEAD award,” says Yanshan Wang, PittNAIL director and Department of Health Information Management vice chair for research and assistant professor.
“This recognition reflects the strength of the clinical research informatics and AI capabilities we have built at the PittNAIL lab, as well as our commitment to training the next generation of professionals and researchers who can translate interoperable health data into real-world impact.”
– Yanshan Wang
What is the problem that this work will address?
Artificial intelligence is rapidly transforming medicine, promising breakthroughs in clinical decision support, research discovery and personalized care. Yet one of the most significant barriers to realizing AI’s full potential in health care lies not in algorithms, but in data.
Real-world electronic health record data are often fragmented, messy and inconsistently structured across systems, making them difficult to share, analyze and reuse at scale. These challenges severely limit the development, validation and deployment of robust medical AI solutions.
FHIR has emerged as a critical standard designed to address these issues by enabling standardized, interoperable access to health data. Despite its transformative potential, however, FHIR adoption and skilled use remain limited across clinical, research and technology settings—creating an urgent need for a new generation of experts who can bridge health care data and AI innovation.
This mission directly aligns with our ongoing efforts at SHRS and in PittNAIL to build ReDWINE (Rehabilitation Datamart with Informatics Infrastructure for Research), which leverages AI and large language models to transform clinical data from hospitals into a standardized format that enables researchers to easily access and analyze it. This is exciting because it enables SHRS researchers across various specialties to request data for their studies without worrying about the complexities of different data sources.
For example, imagine a researcher studying how patients recover after knee replacement surgery. With ReDWINE, they can easily access surgery records, rehabilitation progress and follow-up outcomes in one place, helping them identify which rehabilitation approaches deliver the best patient outcomes.
As rehabilitation research increasingly relies on real-world clinical data, expanding our workforce with strong FHIR expertise is critical to enabling efficient data integration, reuse and advanced analytics.
How does this fellowship address the work that needs to be done?
“Throughout my career, I have built expertise in AI and machine learning, but I recognized that I needed a deeper understanding of health data standards to make a real impact in clinical research. This training program is an excellent opportunity to bridge that gap and bring these skills directly to the ReDWINE project, where we are working to make clinical data more accessible for rehabilitation research.”
– Min Gu Kwak
This AIM-AHEAD award training program strengthens our ability to turn real-world clinical data into meaningful improvements in patient care. The skills gained will directly enhance ReDWINE, our centralized rehabilitation data warehouse, by enabling clinical data to be accessed, standardized and reused more efficiently by rehabilitation researchers at SHRS. Just as importantly, FHIR makes it possible to translate research findings back into electronic health care record systems, closing the loop between research and clinical practice.
For the public, this means that data generated during routine health care visits—such as physical therapy sessions, hospital stays or post-surgical follow-ups—can be responsibly leveraged to accelerate research and improve care. When researchers can access high-quality, interoperable data, they can conduct studies more quickly, identify effective rehabilitation strategies sooner and help clinicians make more informed decisions.
Ultimately, this work bridges the gap between hospital data and real-world impact, leading to better recovery outcomes for patients with conditions such as stroke, orthopedic injuries and other rehabilitation needs.
Can you also explain what FHIR are and their importance to the public?
FHIR is a technical standard that enables different health care computer systems to communicate using modern web technologies. Think of it as an “internet for health care.”
Historically, health records were trapped in silos because systems used incompatible formats and outdated exchange methods. FHIR solves this by providing two things: a common data format (standardizing how a “patient” or “medication” is defined) and an API (application programming interface) that allows systems to request and share specific data instantly—much like how a travel app pulls flight data from different airlines.
For patients, this creates a seamless experience where medical history, test results and medications follow them from provider to provider, enabling better-coordinated care and allowing them to access their own data on mobile apps. For researchers, this allows for the rapid aggregation of high-quality real-world data. Instead of spending months cleaning messy, inconsistent data from different hospitals, researchers can pull standardized datasets across institutions. This accelerates clinical trials, population health studies and the training of AI models.
What does this second AIM-AHEAD award mean for PittNAIL?
In our lab, we believe that extraordinary impact comes from empowering every PittNAILer, from undergraduates to PhD trainees, staff scientists and engineers, to dream bigger and step beyond their comfort zones. We encourage every trainee to pursue opportunities that stretch their skills and ambitions, because growth only happens when you are willing to try, persist and learn from the process. This second AIM-AHEAD award is a powerful testament to the resilience and curiosity of our trainees, and to a culture that celebrates effort as much as achievement.
More broadly, this recognition affirms the vision we are building at PittNAIL: a nationally visible hub for informatics and AI that advances medicine while remaining grounded in ethics, fairness and rigorous evaluation. Our goal is not just to develop new technologies, but to train the next generation of leaders who will shape how AI is responsibly integrated into health care. Seeing our lab members recognized at the national level gives me great confidence that we are on the right path and that their work will ultimately translate into better, more equitable care for patients.
Read More
Health Informatics Graduate Awarded NIH Training Fellowship to Increase Diversity in Artificial Intelligence and Machine Learning
Master of Science in Health Informatics alumnus Jordan Hilsman (MSHI ‘23) has been selected among a competitive pool of nationwide candidates for an eight-month National Institutes of Health training program called AIM-AHEAD to increase researcher diversity in artificial intelligence and machine…
Read the full story
