Current Scholars
Meet this year’s exceptional scholars who are driving implementation of effective strategies to raise awareness of diagnosis in medicine, support diagnostic excellence, and reduce diagnostic errors at the national level. If you would like to connect with any of them, please contact us.
Dania Daye, MD, PhD
University of Wisconsin-Madison
Dr. Daye is an associate professor of radiology, Vice Chair for Practice Transformation and the inaugural Director of the Center of High Value Imaging at the University of Wisconsin (UW)-Madison. Prior to joining UW in 2025, Dr. Daye was an associate professor of radiology at Massachusetts General Hospital (MGH) and Harvard Medical School. While at MGH, she served as associate medical director of the Center of Outcomes and Patient Safety in Surgery, associate chair in the Department of Radiology, and Director of the Precision Interventional and Medical Imaging Lab at the MGH/HST Martinos Center for Biomedical Imaging. She currently serves in a number of leadership positions in national radiology societies, including serving on the board of directors of the SIR Foundation.
Dr. Daye is a board-certified interventional radiologist. Her research centers on the applications of machine learning and computer vision for precision medicine. Dr. Daye’s NAM proposal focuses on building an agentic AI framework for end-to-end closed loop management of incidental findings in radiology. To date, her research has resulted in more than 100 peer-reviewed publications and more than 150 presentations and invited talks at both national and international meetings. For her research, Dr. Daye is the recipient of many awards that include the Association of American Physicians Stanley J. Korsmeyer Young Investigator Award and the Gary Becker Young Investigator Award.
Dr. Daye received a B.S. in bioengineering from Rice University. She is a graduate of the University of Pennsylvania MD-PhD program, completing her Ph.D. in bioengineering as a Howard Hughes Medical Institute-National Institute of Biomedical Imaging and Bioengineering (HHMI-NIBIB) Interfaces Scholar in imaging sciences. Dr. Daye was previously selected as a Howard Hughes Medical Institute Gilliam Fellow and as a Paul and Daisy Soros Fellow.
Proposal: Multi-Agentic AI System for End-to-End Management of Incidental Findings in Diagnostic Imaging
Ahmed Hassoon, MD, MPH
Johns Hopkins University
Dr. Hassoon is an assistant research professor at the Johns Hopkins Bloomberg School of Public Health, with joint appointments in the Department of Neurology at the Johns Hopkins School of Medicine and the Center for Language and Speech Processing at the Whiting School of Engineering. He also serves on the Johns Hopkins Hospital AI Governance Sub-Council and the Bloomberg School’s Public Health + AI Strategic Endeavors committee, shaping how AI is evaluated and deployed across the Johns Hopkins enterprise.
Dr. Hassoon’s work sits at the intersection of clinical medicine, epidemiology, and AI, with a singular focus: making diagnosis safer and reducing the preventable harm that comes from diagnostic error. His research has produced tools that are now shaping national diagnostic safety practice. He led the development of the computable phenotype for the Symptom-Disease Pair Analysis of Diagnostic Error (SPADE) methodology, which has been adopted by the Agency for Healthcare Research and Quality (AHRQ) for national implementation. Through AHRQ’s Tools and Guidance for Diagnostic Excellence, hospitals across the country can now measure diagnostic safety at a system level for the first time, supported by a publicly available Python package he helped build. Dr. Hassoon also developed the first AI benchmark for diagnostic error detection, evaluating 16 state-of-the-art large language models as clinical safety nets across 20 of the most frequently misdiagnosed conditions. The best-performing model corrected 55 percent of physician diagnostic errors. Dr. Hassoon’s current focus is the next frontier in diagnostic safety: bringing the patient, the long-ignored third pillar of evidence-based medicine, into the AI-augmented diagnostic process as an active and informed partner.
Earlier in his career, Dr. Hassoon served as the lead analyst for the State of Qatar National Public Health Strategy 2017 to 2022, which enabled Qatar to become the first entity outside the United States to receive accreditation from the Public Health Accreditation Board. He has been recognized as a Young Physician Leader by the InterAcademy Partnership at the World Health Summit, named among Diplomatic Courier’s “99 Under 33” foreign policy leaders, and received a Certificate of Appreciation from the President of the United States for his work in global health.
Dr. Hassoon’s academic path began with a medical degree from Baghdad College of Medicine, followed by a Hubert Humphrey Fellowship at Tulane University, an M.P.H. in epidemiology from Johns Hopkins, and a postdoctoral fellowship in AI for clinical trials. He is currently completing a Ph.D. in computer science at the Johns Hopkins Center for Language and Speech Processing, where his research focuses on multimodal large language model training, evaluation, and continuous learning for clinical diagnostics. He is the recipient of an AHRQ Mentored Clinical Scientist Research Career Development Award (K08) focused on improving lung cancer diagnostic safety at Johns Hopkins Hospital and a 2024 Fellow of the Institute for Healthcare Improvement in Diagnostic Safety.
Proposal: The Patient Advocate to Empowering Patients in the Diagnostic Process: A Safety Framework and Modus Operandi for Agentic AI-Mediated Error Interception
Shuhan He, MD
Massachusetts General Hospital/Harvard Medical School
Dr. He is an emergency physician and clinical informatician at Massachusetts General Hospital (MGH) and Harvard Medical School, where he holds appointments as an assistant professor of emergency medicine and assistant professor of medicine. Dr. He also serves as Chair of the Emergency Medicine Informatics Section at the American College of Emergency Physicians and as Program Director for the Master of Science in Healthcare Data Analytics at the MGH Institute of Health Professions, where he directs curriculum spanning machine learning, data visualization, data ethics and regulation, and analytic leadership.
Dr. He’s research focuses on diagnostic uncertainty, clinical decision support, and machine learning in emergency medicine. His laboratory published foundational work measuring how much each test and clinical finding reduces a clinician’s diagnostic uncertainty (Scientific Reports, 2024), providing the theoretical basis for a structured approach to the diagnostic process. This research led to the development of diagnosticuncertainty.org, a digital tool that helps clinicians determine when they have reduced uncertainty enough to safely act on a disposition decision. His team has extended this work to evaluate the uncertainty-reducing value of over 400 clinical features and to develop predictive models for emergency department resource allocation and post-discharge readmission risk.
Beyond research, Dr. He has a track record of building digital tools that move from concept to deployed, large-scale impact. During the COVID-19 pandemic, he co-founded GetUsPPE.org, which distributed over 18 million units of personal protective equipment to healthcare facilities nationwide. The platform’s real-time shortage data, published in the Lancet, informed the Biden administration’s national COVID-19 response, and its matching system was published in NPJ Digital Medicine. He authored the anatomical heart and lung emojis for the Unicode Standard, the universal character encoding system that governs text display on every smartphone, computer, and electronic health record, now rendered on over 5 billion devices worldwide. He then applied this work clinically, publishing an emoji-based visual analog scale in JAMA that is now widely used as a digital replacement for the Wong-Baker FACES Pain Rating Scale. He founded ConductScience.com, a research technology company whose instrumentation, machine learning, and computer vision systems, featured in Nature, Popular Science, and Discover Magazine, are now used by over 5,000 researchers at more than 1,200 institutions worldwide and have supported research published in 280 peer-reviewed papers across journals including Cell, Neuron, and eLife. In 2026 alone, ConductScience won prizes in three separate NIH competitions: the NIH S-Index Challenge, for novel metrics to measure and reward data sharing; the NIH Replication Prize, for integrating replication into standard research practice; and Phase 1 of the NIH Office of Dietary Supplements challenge, for turning federal health information into interactive consumer-facing tools. This pattern of translating clinical and technical insight into widely adopted tools directly informs the proposed project: taking a working diagnostic uncertainty prototype and scaling it for national dissemination.
Dr. He completed his M.D. at the University of Southern California Keck School of Medicine with Honors Distinction in Research and his residency at the Harvard Affiliated Emergency Medicine Residency. He has been recognized as a Modern Healthcare Top 25 Emerging Leader (2021), Boston Business Journal 40 Under 40 (2021), and EMRA 25 Under 45 Influencer in Emergency Medicine (2020). Dr. He is dual board-certified in emergency medicine and clinical informatics.
Proposal: Diagnostic Safety by Design: Clinician-Centered Redesign and Multi-Site Deployment of a Digital Uncertainty-Reduction Tool for Emergency Medicine
Julius Oatts MD, MHS
Children’s Hospital of Philadelphia/University of Pennsylvania Perelman School of Medicine
Dr. Oatts is a board-certified pediatric ophthalmologist on faculty at the University of Pennsylvania Perelman School of Medicine and Children’s Hospital of Philadelphia. He is currently the recipient of a National Eye Institute K23 career development award which involves the creation and evaluation of clinic- and community-based childhood vision screening interventions in Nepal and is the first of its kind to compare the diagnostic accuracy of two different amblyopia screening strategies. He was also named a William T. Grant Scholar in 2024 for work developing a population-based study in the Navajo Nation to determine the burden of undetected childhood vision loss and the most effective ways to diagnose eye disease in this setting.
Clinically, Dr. Oatts has expertise in retinopathy of prematurity (ROP), which serves as the foundation for the proposed project. He has evaluated the efficacy of different treatments for ROP as well as assessed the impact of social determinants of health in ROP diagnosis. He has also evaluated national trends in ROP diagnosis and care using data from the Healthcare Cost and Utilization Project Kids’ Inpatient Database. Dr. Oatts was the course director for the 6th biennial ROP Update, an international scientific meeting of clinicians and researchers dedicated to improving the diagnosis and treatment of ROP. Additionally, Dr. Oatts is dedicated to mentoring the next generation of pediatric ophthalmologists and co-founded the Pediatric Ophthalmology Mentorship (POM) program. This mentorship program currently supports over 50 medical students from across the country with an early interest in pediatric ophthalmology. This program is supported by the American Association for Pediatric Ophthalmology and Strabismus (AAPOS) and originated from Dr. Oatts’ participation in the American Academy of Ophthalmology (AAO) Leadership Development Program. Through the POM program, he has identified factors which influence resident pursuit of a career in pediatric ophthalmology fellowship and will continue to monitor the program to support the pediatric ophthalmology career pathway.
Nationally, Dr. Oatts has several leadership roles and serves as Chair of the AAPOS Technology Committee and Vice Chair of the AAPOS Professional Education Committee. These roles involve monitoring and disseminating information about new technology and literature to the pediatric ophthalmology community. He also serves on the AAO Ophthalmic Technology Assessment Committee and Preferred Practice Patterns Committee, both of which are tasked with developing evidence-based assessments and clinical practice guidelines that define high-quality ophthalmic care. Dr. Oatts recently received the AAPOS 2026 Outstanding Emerging Leader Award.
Dr. Oatts attended medical school at Yale School of Medicine, completed ophthalmology residency at University of California, San Francisco, and pediatric ophthalmology fellowship at Harvard Medical School/Boston Children’s Hospital.
Proposal: Improving Retinopathy of Prematurity Diagnostic Verification and Predictive Modeling Using a Novel Image Grading Score
Andrei S. Purysko, MD, FSAR
Cleveland Clinic
Dr. Purysko is a board-certified radiologist and Section Head of Abdominal Imaging at the Cleveland Clinic, where he also serves as an associate professor of radiology at the Cleveland Clinic Lerner College of Medicine.
Dr. Purysko’s work integrates clinical practice, quality improvement, and research, with a focus on translating evidence-based standards into scalable, real-world solutions. He has led multidisciplinary collaborations across radiology, urology, and national organizations to address gaps in imaging quality and their impact on diagnosis and management. His scholarly contributions include peer-reviewed publications on prostate MRI quality, radiomics, and diagnostic performance. Dr. Purysko’s current work centers on advancing standardized image quality assessment using Prostate Imaging–Reporting and Data System (PI-RADS) and evaluating its impact on prostate cancer detection, with the goal of improving diagnostic accuracy and outcomes at scale.
Dr. Purysko has been a national leader in efforts to improve prostate MRI quality. He has played a central role in multiple American College of Radiology (ACR) initiatives, including serving as Physician Leader of the Prostate MR Image Quality Improvement Collaborative and Chair of the ACR Prostate Cancer MRI Center Designation Subcommittee. He is also an active member of the PI-RADS Steering Committee, where he chairs the Quality and Practice Improvement Subcommittee. Through these roles, he has contributed to the development, implementation, and dissemination of standards aimed at improving prostate MRI acquisition and interpretation.
Dr. Purysko received his M.D. from Faculdade de Medicina de Petrópolis in Rio de Janeiro, Brazil. He completed internal medicine residency at Hospital do Servidor Público Estadual de São Paulo, diagnostic radiology residency at Hospital Beneficência Portuguesa de São Paulo, and abdominal imaging and nuclear medicine fellowship at Cleveland Clinic.
Proposal: Improving Prostate Cancer Diagnosis Through Standardized Prostate MRI Image Quality Assessment
Adam Rodman, MD, MPH
Beth Israel Deaconess Medical Center
Dr. Rodman is a general internist and medical educator at Beth Israel Deaconess Medical Center (BIDMC) and an assistant professor of medicine at Harvard Medical School. He is the Director of AI Programs at the Carl J. Shapiro Institute for Research and Education at BIDMC, and he leads the steering group for integration of AI into the medical school curriculum. He is also a visiting researcher at Google DeepMind, where he works on oversight systems for patient-facing AI. He is also an Associate Editor at NEJM AI.
Dr. Rodman’s research focuses on clinical reasoning, medical education, and the integration of AI into triadic (patient-AI-physician) care models. He co-founded the ARISE research network with Jonathan Chen at Stanford and has served as principal investigator on multiple randomized controlled trials, including the first pre-registered clinical trial of a patient-facing AI system. He also co-founded the iMED Initiative with Shreya Trivedi to study the impacts of digital technologies on medical learning. His work is currently supported by the NIH, ARPA-H, the Gordon and Betty Moore Foundation, the Macy Foundation, and Google Research.
Dr. Rodman is the author of Short Cuts: Medicine, a featured instructor in MasterClass, and hosts the American College of Physicians podcast Bedside Rounds. His research and commentary are frequently covered by international news media.
Dr. Rodman received his M.D. and his M.P.H. from Tulane University. He completed his residency in internal medicine at Oregon Health and Science University and his global health fellowship at Beth Israel Deaconess Medical Center, which included clinical work in Molepolole, Botswana.
Proposal: Laying the Groundwork for an Effective, Evidence-Grounded LLM-Based Second Opinion Trigger System for High-Risk Hospitalized Inpatients
Cory Rohlfsen, MD
University of Nebraska Medical Center
Dr. Rohlfsen is an associate professor in the Department of Internal Medicine at the University of Nebraska Medical Center (UNMC). As a primary care physician, hospitalist educator, and core faculty member of the Internal Medicine residency at UNMC, his scholarship focuses on making diagnostic reasoning measurable, reproducible, and coachable at scale.
Dr. Rohlfsen’s team has developed an AI-enabled platform that analyzes clinical interviews to assess how efficiently learners gather and prioritize information when working toward a diagnosis. Through multi-institutional research, his team has demonstrated that learners who reach the correct diagnosis ask more focused, hypothesis-driven questions than those who do not. These findings suggest that the efficiency of information gathering likely serves as a measurable construct of diagnostic excellence.
This work builds on a broader record of innovation in faculty development. As founder and director of the Health Educators and Academic Leaders (HEAL) program, Dr. Rohlfsen established a competency-based, interdisciplinary clinician-educator track integrating observed structured teaching encounters (OSTEs) across 19 disciplines. He is involved in mentoring residents in the Primary Care Program and is passionate about fostering a community of teaching excellence for future academicians. As a NAM Scholar in Diagnostic Excellence, Dr. Rohlfsen seeks to advance practical, scalable methods for AI-augmented simulation to strengthen diagnostic performance across training programs and health systems. He aims to collaborate with national leaders to refine information-theoretic approaches against established assessment methods (e.g., OSCE rubrics) to improve diagnostic accuracy, simulate uncertainty management, and measurably improve patient care.
Dr. Rohlfsen received his M.D. from UNMC. He completed his internal medicine residency at UNMC.
Proposal: Can Diagnostic Excellence Be Measured?
Lucy Schulson, MD, MPH
Boston University Chobanian & Avedisian School of Medicine/Immigrant and Refugee Health Center
Dr. Schulson is a clinician investigator and implementation scientist dedicated to advancing diagnostic equity and patient safety in ambulatory care settings. She is an assistant professor of medicine at the Boston University Chobanian & Avedisian School of Medicine and an attending physician in the Section of General Internal Medicine at Boston Medical Center, New England’s largest safety-net hospital. She is also a practicing primary care physician in the Immigrant and Refugee Health Center.
Dr. Schulson is the Principal Investigator of an American Heart Association Career Development Award (2025-2028) focused on diagnostic equity in heart failure with preserved ejection fraction—identifying opportunities for earlier diagnosis to reduce disparities in care. Her research focuses on system-level contributors to diagnostic delays and missed diagnoses, with particular attention to how fragmented care, communication barriers, limited English proficiency, and restricted access disproportionately affect populations that have been marginalized. She is a faculty fellow in Boston University’s Center for Implementation and Improvement Sciences and brings expertise in health services research, mixed methods, and implementation and improvement sciences.
From 2020 to 2023, Dr. Schulson served as Associate Physician Policy Researcher at RAND Corporation, where she contributed to interdisciplinary, policy-relevant studies examining patient safety, health system performance, quality of care, and the intersection of racism and patient safety. Her research has been published in leading peer-reviewed journals including JAMA Network Open, BMJ Quality & Safety, and the Joint Commission Journal on Quality and Patient Safety.
Dr. Schulson earned her B.A. in international development from Brown University, her M.D. from the Icahn School of Medicine at Mount Sinai (where she graduated with distinction in Medical Education and Research and was elected to the Alpha Omega Alpha honor society), and her M.P.H. in health policy and management from the Harvard School of Public Health. She completed her internal medicine residency at Beth Israel Deaconess Medical Center at Harvard Medical School where she was part of the primary care track and received the Stoneman Center Resident Award for Quality Improvement and Patient Safety.
Proposal: Bridging Diagnostic Gaps: Using AI to Identify Missed Heart Failure with Preserved Ejection Fraction Diagnosis to Improve Diagnostic Equity in Ambulatory Care
Kathleen E. Walsh, MD, MSc
Boston Children’s Hospital
Dr. Walsh is a practicing primary care pediatrician and nationally recognized expert in pediatric medication safety and implementation science. She currently serves as the Director of Healthcare Quality and Safety Research Program at Boston Children’s Hospital. She also directs the Harvard-wide Pediatric Health Services Research Fellowship, the Boston Children’s Health Equity Fellowship, and the Healthcare Safety and Quality Research program in the Division of General Pediatrics. In addition, she leads the only pediatric Diagnostic Center of Excellence, focusing on diagnostic excellence and health equity in the ambulatory setting.
For over 20 years, Dr. Walsh has developed and validated innovative research methods to study ambulatory medication safety. She has conducted multi-method research including multisite descriptive and mixed methods research, intervention development and evaluation, implementation studies, studies with EHR-derived data, pharmacoepidemiologic studies, and several studies embedded within health systems or learning networks. Dr. Walsh has also advanced training in quality improvement methods and has expertise in context and implementation factors associated with the spread of interventions, including a current randomized trial (R18 HS027401). Her research has been funded by the NCI, NIH, FDA, PCORI, the Robert Wood Johnson Foundation, and AHRQ (1 R01 and 3 R18).
Dr. Walsh’s current research program focuses on empowering families to optimize care and outcomes for children with chronic illness in the home setting. Her proposal for the NAM Scholars program seeks to improve communication between clinicians and families and develop a shared mental model when there is uncertainty around a child’s diagnosis.
Dr. Walsh received her B.A. from Cornell University and her M.D. from Georgetown University. She completed her pediatrics residency at the Rhode Island Hospital and her general pediatrics fellowship at Boston University School of Medicine. She also holds an M.Sc. in epidemiology from Boston University School of Public Health.
Proposal: Improving Communication of Pediatric Diagnostic Uncertainty with Outpatient Families
Yize Zhao, PhD
Yale School of Public Health
Dr. Yize Zhao is a tenured associate professor of biostatistics at the Yale School of Public Health, with a secondary appointment in Biomedical Informatics and Data Science at Yale School of Medicine. She is affiliated with the Alzheimer’s Disease Research Center, the Wu Tsai Institute, the Center for Brain and Mind Health, the Center for Analytical Sciences, the Computational Biology and Bioinformatics Program, the Institute for Foundations of Data Science, and the Biomedical Imaging Institute at Yale University.
Dr. Zhao develops rigorous statistical and AI methods for large-scale, complex biomedical and health data, with applications in aging, neurodegenerative disease, psychiatry, and mental health. Her broader goal is to translate methodological innovation into clinically meaningful, trustworthy tools for diagnostic support and risk assessment. She has published approximately 100 peer-reviewed papers in leading statistical, biomedical, and AI venues, including senior-authored work in Nature Methods, Nature Communications, JAMA Psychiatry, Alzheimer’s & Dementia, and the Journal of the American Statistical Association. She is the principal investigator of multiple NIH-funded R01 projects in translational Alzheimer’s and mental health research and provides analytical leadership across interdisciplinary initiatives and research centers in aging, psychiatry, and biomedical data science. Through her research, leadership, and mentoring, Dr. Zhao is committed to developing analytically rigorous and implementation-ready approaches that improve clinical decision-making in real-world care. Her proposed NAM Scholars program reflects that mission, advancing equitable, uncertainty-aware, and clinically actionable tools for improving diagnosis in older adults.
Dr. Zhao’s contributions have been recognized with prestigious mid-career honors, including the COPSS Emerging Leader Award from the Committee of Presidents of Statistical Societies and the IMS Thelma and Marvin Zelen Emerging Women Leaders in Data Science Award from the Institute of Mathematical Statistics. She holds Associate Editor appointments at several leading journals, including the Journal of the American Statistical Association and Biometrics, the flagship journals of her field. She is also a standing member of the NIH Biodata Management and Analysis Study Section.
Dr. Zhao completed her Ph.D. in biostatistics at Emory University and her postdoctoral fellowship at the Statistical and Applied Mathematical Sciences Institute/National Institute of Statistical Sciences, joint with the Department of Biostatistics at the University of North Carolina-Chapel Hill.
Proposal: Dynamic Diagnostic Surveillance: An Equity-Centered, Uncertainty-Aware Framework for Timely ADRD Diagnosis from EHRs