PhD Student Uses AI and Machine Learning to Advance Nursing Research
At the College of Nursing, PhD students are using artificial intelligence, machine learning, and data science to investigate complex health care questions and develop novel approaches to nursing research.
Todd Adams, MSN, RN, PMHNP-BC, is one of those students. Adams used machine learning as part of a six-person interdisciplinary research team studying mortality risk among adults with substance use disorders.
Adams participated in the research through AIM-AHEAD, an eight-month artificial intelligence and machine learning training program sponsored by the National Institutes of Health (NIH) and the National Center for Advancing Translational Sciences (NCATS). The program provides researchers with AI and machine learning training, mentorship, and hands-on experience with large-scale clinical data.
The team's work received the 2026 AIM-AHEAD Health Data Science Training Program's Top AI/ML Translational Research Award, Instructor's Choice. The team presented its research at the July AIM-AHEAD meeting in San Diego and was notified of the award in August.
For the project, the researchers used unsupervised machine learning, a form of artificial intelligence that enables computers to identify patterns in data without researchers specifying them in advance. They analyzed a large national collection of electronic health records to identify groups of adults with substance use disorders who shared similar combinations of health conditions.
These groups, known as multimorbidity phenotypes, reveal how multiple health conditions co-occur. The researchers then compared the groups to determine whether they had different rates of adjusted all-cause mortality, or the risk of death from any cause after accounting for differences among the groups.
One group stood out: younger patients with a behavioral health-driven pattern of multiple co-occurring conditions had a higher risk of death than their age alone would predict.
The finding suggests that risk among people with substance use disorders cannot always be explained by age or by use of a single substance. Considering a patient's full combination of health conditions may provide a clearer picture of their risk.
The team used k-modes clustering, an unsupervised machine learning method that groups patients by similarities in their health conditions. The analysis was conducted using national electronic health record data in the secure National Clinical Cohort Collaborative (N3C) Data Enclave, a protected environment that enables researchers to study large volumes of clinical data while maintaining strict data security and privacy.
For Adams, the project offered an opportunity to integrate nursing science, data science, and machine learning in a real-world research setting. “This project was right in the middle of what I want to do as a nurse scientist: using data science and machine learning to make complex clinical data more useful,” Adams said. “It gave me the chance to do that work with a large real-world dataset and a strong interdisciplinary team.”
Adams was among 50 clinicians and researchers selected nationally for the 2025–26 AIM-AHEAD Machine Learning Training Program. Through the program, he joined researchers from the University of Arizona, the University of Texas at Austin, Albert Einstein College of Medicine, Duke, Yale, and Rutgers for the interdisciplinary project, with Hadis Hashemi of UT Austin serving as mentor.
Working with researchers from different institutions and areas of expertise gave Adams experience beyond learning new technical skills. The team made methodological decisions, evaluated their analyses, interpreted their findings, and communicated their work to researchers with different perspectives.
“AIM-AHEAD was one of the most valuable learning experiences I’ve had during my PhD program,” Adams said. “We worked with a large clinical dataset, made real methodological decisions, and explained our findings to people across several disciplines. The award was a nice surprise, but the experience itself was the bigger win for me.”
Adams, a College of Nursing PhD student since 2023, is preparing for comprehensive exams and moving toward dissertation work. Through the project, he gained experience in computational phenotyping, using data to identify meaningful patient groups; unsupervised machine learning; and cluster evaluation, which helps researchers determine whether those groups represent meaningful patterns. He will carry these skills into his dissertation and future research and hopes the team's work will lead to a publication and continued collaboration.
Adams credits Sheila Gephart, PhD, RN, FAAN, interim associate dean for academic and faculty affairs and professor at the College of Nursing, with bringing the AIM-AHEAD opportunity to his attention and encouraging him to pursue it. Gephart also serves as Adams' PhD advisor and committee chair.
“Todd has embraced the opportunity to explore how emerging data science and machine learning approaches can advance nursing research,” said Gephart. “His work through AIM-AHEAD reflects both his commitment to developing new research skills and his ability to collaborate across disciplines to address complex health care questions. I am excited to see how he builds on this experience as he advances his dissertation and career as a nurse scientist.”
As Adams continues to develop as a nurse data scientist, his experience illustrates how doctoral education at the College of Nursing can intersect with emerging fields such as artificial intelligence, machine learning, and data science.
As these technologies become increasingly important in health care research and patient care, nurse scientists provide a critical clinical perspective to interdisciplinary teams. Their understanding of patient care helps ensure that emerging technologies are applied to meaningful health care problems.
Adams' experience illustrates the opportunities for College of Nursing PhD students to engage with emerging technologies and collaborate across disciplines, thereby developing the skills to advance nursing research and shape the future of AI-enabled health care.