CPHAI Symposium 2025

AI in Healthcare: From Research to Real-World Clinical Impact

HANOVER INN, GRAND BALLROOM

Friday, APRIL 18th

CPHAI Symposium 2025 Overview

Join us for the Annual Dartmouth Symposium on Precision Health & AI on April 18, 2025, at the Hanover Inn Grand Ballroom. This year’s symposium brings together leaders in AI, clinicians, and biomedical researchers to explore the latest advancements in AI-driven healthcare.


Keynote speakers include Dr. Pranav Rajpurkar (Harvard University, Forbes 30 Under 30, MIT Tech Review Innovator Under 35) and Dr. Nigam Shah (Stanford University, Chief Data Scientist for Stanford Health Care), who will share their expertise on AI in medical imaging, clinical data interpretation, and ethical AI deployment in healthcare. The event will open with remarks by Dr. Duane Compton, the Dean of Dartmouth Geisel School of Medicine and include discussions on AI leadership, research frontiers, and industry collaborations.

Keynote Presentations

Pranav Rajpurkar, PhD

Harvard University

Beyond Assistance: Rethinking AI-Human Integration in Radiology

Recent evidence challenges a fundamental assumption in medical AI: that combining AI with physician expertise naturally leads to better outcomes. Studies show that AI assistance often fails to improve diagnostic accuracy and can even slow down clinical workflows. This talk presents an alternative vision: instead of forcing integration, we should embrace clear role separation between AI systems and physicians. Drawing from recent large-scale studies and advances in generalist medical AI systems, I will examine promising models where AI and doctors work separately but complementarily, each leveraging their unique strengths. Through practical examples and emerging evidence, I will demonstrate how this approach could transform clinical practice while maintaining the essential role of human medical expertise.

  • Dr. Pranav Rajpurkar is Assistant Professor of Biomedical Informatics at Harvard Medical School and Co-founder of a2z Radiology AI. A Stanford-trained computer scientist, his research focuses on developing AI systems that match physician-level expertise through multi-modal clinical data interpretation. At Harvard, his lab develops frameworks for evaluating and advancing medical AI in clinical language models and medical imaging. His research includes over 100 publications in journals like Nature, NEJM, and Nature Medicine, garnering over 35,000 citations. Prof. Rajpurkar has educated over 84,000 students through his AI in Medicine courses at Harvard and Coursera. Named among MIT Tech Review's Innovators Under 35 (2023), Forbes 30 Under 30 in Science (2022), and Nature Medicine's Early-career Researchers To Watch (2022), he continues to advance the integration of AI in clinical practice.

Nigam H. Shah, MBBS, PhD

Stanford University

Responsible AI in a Healthcare System

As the use of artificial intelligence (AI) moves from being a curiosity to a necessity, it is clear that the benefit obtained from using AI models to prioritize care interventions is an interplay of the model’s performance, the capacity to intervene, and the benefit/harm profile of the intervention. After a brief review of the kinds of use cases that AI can serve across multiple medical specialties, we will discuss Stanford Healthcare’s efforts to shape the adoption of health AI tools to be useful, reliable, and fair so that they lead to cost-effective and sustainable solutions. The conversation will draw on examples from multiple specialities, including Pathology, Cardiology, Internal Medicine, Surgery, Psychiatry and Oncology, as well as discuss the implications of the choice of business model to ensure the use of AI enhances care quality while managing healthcare costs. We will discuss how the adoption of LLMs in medicine needs to be shaped by performing the evaluations that specify the desired benefits and verifying those benefits via testing in real-world deployments. We will conclude with the rationale and vision for collaborative activities such as the Coalition for Health AI (CHAI) and the Medical Event Data Standard (MEDS)

  • Dr. Nigam Shah, Professor of Medicine and Biomedical Data Science at Stanford, Associate Director of the Stanford Center for Biomedical Informatics Research, Co-Director of the Stanford Center for Artificial Intelligence in Medicine and Imaging, and Associate Dean for Research and Chief Data Scientist at Stanford Health Care. Dr. Shah's work focuses on AI's ethical and practical implementation in healthcare. His research aims to bring AI into clinical use safely, ethically, and cost-effectively. Dr. Shah is an inventor on eight patents, has authored over 300 scientific publications, and has co-founded three companies. He was inducted into the American College of Medical Informatics in 2015 and the American Society for Clinical Investigation in 2016. He holds an MBBS from Baroda Medical College, India, a PhD from Penn State University and completed postdoctoral training at Stanford University.

Opening Remarks

Duane Compton, PhD

Dean of the Geisel School of Medicine at Dartmouth

Guest Speakers

Pedram Hosseini, PhD

AI Lead Scientist at Lavita.AI

Title: Toward Meaningful Progress in Clinical AI

High-quality evaluation is a critical component of building any robust AI product or pipeline. This becomes even more essential in a highly sensitive domain like healthcare, where deploying AI models in real-world clinical settings directly impacts patients' health and well-being. Currently, there is a significant gap between existing tools and benchmarks for rigorously testing AI models in healthcare and the need for thorough evaluations in real-world medical and clinical tasks. In this talk, we will review the current state of AI model evaluation in the medical domain, highlighting key use cases such as medical AI assistants and medical question-answering systems. We will then discuss the steps we are taking—through partnerships with academic and industry labs—to bridge this gap, including a benchmark we developed from consumer medical questions on Lavita's Medical AI Assist platform, in collaboration with medical professionals at Dartmouth.

  • Pedram Hosseini leads AI initiatives at Lavita, a healthcare AI venture focused on building a healthcare data platform to facilitate access to personal health data, advance biomedical research, and improve healthcare. In addition to his work at Lavita, Pedram is a Senior Associate at Camford Capital, a venture capital firm in Menlo Park, where he collaborates with an interdisciplinary team on AI-driven innovations and venture incubation across multiple verticals—Lavita AI being one example. His role spans both research and investment, bridging the gap between cutting-edge AI and real-world impact while also helping to identify promising early-stage ventures and founders.

     

    Pedram earned his PhD in Computer Science from The George Washington University, where his research focused on natural language understanding and the dynamics of how information, including misinformation on health-related topics, propagates online.

Katharina Schmolly, MD

Primary Care Resident Physician at Dartmouth Health & Founder at ZebraMD

Title: Innovation Highlight at Dartmouth Health: AI for Clinical Impact 

Despite the abundance of clinical research, disease databases, and -omics data, integrating this information into real-time clinical practice remains difficult. With 1 in 10 individuals developing a rare disease in their lifetime—on par with diabetes—demand for specialists continues to outpace availability. zebraMD is an AI-powered clinical decision support tool designed to address this critical gap. It aggregates multimodal patient data from electronic health records (EHRs), patient portals, wearables, patient-submitted PDFs, videos, and images to enable predictive modeling and precision management algorithms. The tool functions through a patient-facing app, a standalone provider interface, or direct integration with EHR systems, allowing synchronous or asynchronous AI-driven analysis. zebraMD provides rare disease likelihood scores, suggesting appropriate next diagnostic steps, and, for diagnosed patients, offers evidence-based treatment recommendations through best-practice alerts based on the latest clinical research and guidelines.

  • Dr. Kat Schmolly is a Primary Care Track Internal Medicine Resident at DHMC and a U.S. Air Force Veteran, having served as a flight medic in Aeromedical Evacuation. She earned her medical degree from UCLA David Geffen School of Medicine, where she was recruited through the Leaders of Tomorrow scholarship.

    Before entering medicine, Dr. Schmolly worked as a Cardiac Technician at Medtronic, where she contributed to the development of diagnostic machine learning algorithms and the user interface for the LINQ 2.0 implantable cardiac device. She is also the Founder of ZebraMD, Inc., an AI-driven platform focused on improving diagnostic delays and management in rare and genetic diseases.

    As an Acute Hepatic Porphyria specialist, Dr. Schmolly has a strong interest in applying AI and precision medicine to improve patient outcomes. She plans to pursue a fellowship in Genetics and Preventive Medicine, with the goal of enabling precision medicine at a population-level scale in clinical practice.

Research Panel

This panel brought together leading voices from across Dartmouth and Dartmouth Health to discuss how AI is reshaping the landscape of healthcare research and practice. Dr. Tim Burdick emphasized the promise of predictive analytics and AI integration in clinical workflows, while Dr. David Naeger highlighted how radiology is evolving with AI-assisted tools that enhance, rather than replace, human interpretation. Dr. Margaret Karagas spoke to the importance of maintaining rigorous epidemiological standards and data quality in AI studies, especially in population health. Dr. Parth Shah shared how genomic data is enabling personalized treatment plans, with AI helping unlock complex patterns in hematologic disorders. Dr. Soroush Vosoughi discussed cutting-edge applications in natural language processing and multimodal data fusion to improve patient care and medical documentation. Moderated by Dr. Indrani Bhattacharya, the discussion underscored the importance of collaboration across disciplines, thoughtful implementation, and ethical considerations to ensure that AI serves both clinicians and patients equitably.

Timothy Burdick, MD, MBA, MSc

Associate Chief Research Officer for Informatics at Dartmouth Health

  • Dr. Burdick is Associate Professor and Vice Chair of Research, Community and Family Medicine at Dartmouth Health. He also has secondary appointments in Biomedical Data Science and The Dartmouth Institute for Health Policy and Clinical Practice. He also serves as the DH Associate Chief Research Officer for Informatics. Prior to coming to DH in 2016, Dr. Burdick was the Chief Clinical Research Informatics Officer at Oregon Health and Science University (OHSU) and Director of the Informatics Core of the OHSU CTSA. He started his academic informatics career at the University of Vermont in 2009 and went on to be the Physician Informatics leader for ambulatory and in-patient services. Dr. Burdick’s research focuses on applied clinical informatics, including EHR-based clinical decision support systems to improve healthcare outcomes and delivery of services. He is currently co-Director of the Informatics Core of a P30 grant (Learning Health Systems; mPI: A Tosteson, G Elwyn, T Foster) and P20 COBRE (Rural Health Services Delivery Science, mPI: M Creager, S Wong). He has co-authored more than 25 publications on clinical and research informatics. Dr. Burdick enjoys time with his family, staying active in the outdoors, and participating in triathlons for fun.

David Naeger, MD, FACR, FAAR

Chair of Radiology at Dartmouth Health

  • Dr. Naeger serves as Professor and Chair of the Department of Radiology at Dartmouth Health and the Geisel School of Medicine at Dartmouth. Dr. Naeger earned his medical degree at Duke University, where he graduated AOA. After medical school, he moved to the University of California-San Francisco (UCSF) to complete his residency in Diagnostic Radiology (including serving as a chief resident), a T32 research fellowship, and clinical fellowships in Thoracic Imaging and Nuclear Medicine. After his training, he was appointed to the faculty at UCSF, ultimately serving as the Radiology Department’s Associate Chair for Education. In 2019, Dr. Naeger joined Denver Health as the Director of Radiology overseeing the health system’s team of radiologists and the full complement of 135 radiology technical staff. While in Colorado, Dr. Naeger also served as a Professor and Vice-Chair at Denver Health’s academic affiliate, the University of Colorado, until 2024 when Dr. Naeger was recruited to Dartmouth to serve as Professor and Chair.

Margaret Karagas, PhD

Chair of Epidemiology at Dartmouth College

  • Margaret R. Karagas, Ph.D., is the James W. Squires Professor and founding chair of the Department of Epidemiology at the Geisel School of Medicine at Dartmouth College and Director of its Center for Molecular Epidemiology. She also co-leads a NCI postdoctoral training program to cross-train the next generation of biomedical scientists in bioinformatics, biostatistics and epidemiology which advanced the conceptual framework for the Quantitative Biomedical Sciences graduate program at Dartmouth. Her research seeks to identify emerging environmental exposures, host factors, and mechanisms that affect health from infancy to adult life, and to apply novel technologies and methods, i.e., AI approaches, to understand disease pathogenesis.  She is further committed to translating her findings to policy and practice change. Among her current investigations is a rural cohort study of pregnant women and their offspring in northern New England designed to understand the early life exposome and its impact on lifelong health outcomes. Additionally, her broad range of collaborative studies incorporate novel biomarkers of exposure, individual susceptibility, and biological response to environmental agents including the developing microbiome and immune response. She engages in consortium studies including the Environmental Influences on Child Health Outcomes (ECHO) study, is on the organizing committee of the upcoming NIH Exposome Moonshot Forum and serves on international consensus panels (e.g., for the WHO International Agency for Research on Cancer Monograph Program and European Food Safety) and expert committees (e.g., for U.S. NIH, EPA and NASEM). She received her Ph.D. in epidemiology from the University of Washington.

Parth Shah, MD

  • Parth Shah, MD is a physician-scientist and the Director of Genome Informatics at Dartmouth Hitchcock Medical Center. He spearheaded the creation of the Dartmouth Cloud, a HIPAA-compliant platform designed to support both clinical workloads and research development. He also co-led the design and implementation of a clinical genomics informatics suite for somatic exome and transcriptome sequencing, covering all solid and hematological cancers. Additionally, he serves as an Assistant Professor at the Geisel School of Medicine at Dartmouth and holds board certifications in Hematology, Medical Oncology, and Internal Medicine.

Director of Genome Informatics, Hematology, Dartmouth Health

Soroush Vosoughi, PhD

Assistant Professor of Computer Science, Dartmouth College

  • Prof. Soroush Vosoughi leads the Minds, Machine, and Society group at the Department of Computer Science at Dartmouth College, focusing on natural language processing (NLP) and machine learning. His work critically explores large language models (LLMs) that power tools like ChatGPT, aiming to mitigate their anti-social tendencies and enhance transparency through interpretability methods and reinforcement learning. Vosoughi's team also investigates computational tools for understanding social phenomena, integrating visual information into LLMs, and applying these models to health and bioinformatics. His research, recognized with awards including the Google Research Scholar Award (2022) and an Amazon Research Award (2019), spans collaborations across Dartmouth's faculties and has garnered support from NSF, NIH, and industry giants. Before Dartmouth, Vosoughi completed his academic journey at MIT and contributed significantly to public discourse and health dialogue through technical advisories. 

Indrani Bhattacharya, PhD

Moderator

CPHAI Investigator and Assistant Professor of Biomedical Data Science at Dartmouth College

  • Dr. Bhattacharya earned her bachelor’s degree in electrical engineering from Jadavpur University, India, and her MS and PhD from the Department of Electrical, Computer, and Systems Engineering at Rensselaer Polytechnic Institute, NY. Following her PhD, she joined Stanford University School of Medicine’s Department of Radiology, where she made significant advances in biomedical imaging informatics. Her academic excellence has been recognized with multiple awards, including the RPI’s Founder’s Award of Excellence and being named a Rising Star in EECS by UC Berkeley in 2020. Beyond her professional achievements, Dr. Bhattacharya enjoys traveling, painting, and participating in cultural events, as well as cooking and spending time with friends, enriching her life with experiences that inform and inspire her work.

Leadership Panel

This panel brought together senior academic and clinical leaders to discuss Dartmouth’s evolving vision for AI in healthcare. Panelists emphasized the importance of aligning AI research with patient outcomes, institutional infrastructure, and responsible innovation. Topics included the integration of AI across clinical departments, strategies for fostering cross-disciplinary collaboration, and the challenge of balancing innovation with equity, privacy, and regulatory standards. The discussion highlighted Dartmouth’s strengths in translational science, biomedical data, and patient care, while identifying key opportunities to expand industry partnerships, enhance education pipelines, and build scalable systems that serve both urban and rural populations. Panelists also underscored the value of iterative implementation—developing AI solutions through pilot programs with continuous feedback from clinicians and patients. This session underscored the institution’s commitment to leading AI innovation responsibly and equitably across the academic medical enterprise.

David Kotz, PhD

  • David Kotz is the Provost at Dartmouth, and previously served as Associate Dean of the Faculty for the Sciences, as a Core Director at the Center for Technology and Behavioral Health, and as the Executive Director of the Institute for Security Technology Studies. As the Pat and John Rosenwald Professor in the Department of Computer Science, his current research involves security and privacy in smart homes and wireless networks. He has published over 270 refereed papers, obtained $89m in grant funding, given over 200 invited lectures, and mentored over 100 research students and postdocs. He is an ACM Fellow, an IEEE Fellow, a 2008 Fulbright Fellow to India, a 2019 Visiting Professor at ETH Zürich, and an elected member of Phi Beta Kappa. He received his AB in Computer Science and Physics from Dartmouth in 1986, and his PhD in Computer Science from Duke University in 1991.

Dartmouth Provost

Steven Bernstein, MD

Chief Research Officer at Dartmouth Health

  • Dr. Bernstein is the inaugural Chief Research Officer at Dartmouth Hitchcock Medical Center and the Senior Associate Dean for Clinical and Translational Research at the Geisel School of Medicine at Dartmouth, Director of SYNERGY, Dartmouth’s Clinical and Translational Science Institute, Director of the C. Everett Koop Institute, and Professor of Emergency Medicine and Health Policy and Clinical Practice and at Geisel. His interests are in the use of implementation science methods to expand access to treatment of substance use, development of novel clinical trial designs, and training the next generation of investigators.

    Dr. Bernstein pioneered the treatment of tobacco dependence in emergency department (ED) patients. He was the first to identify specific pharmacologic and behavioral interventions for tobacco dependence that are efficacious in ED settings. His current research focuses on optimizing the treatment of tobacco dependence in individuals with HIV/AIDS. He has additional interests in implementation science and clinical trial design and developed novel methods of assessing stakeholder engagement in implementation research and in measuring patterns of adherence and abstinence in human health behavior.

Susan Roberts, PhD

Senior Associate Dean of Foundational Research at Geisel School of Medicine

  • Susan B. Roberts, PhD, is Senior Associate Dean for Foundational Research, Professor of Medicine and Professor of Epidemiology, Dartmouth College Geisel School of Medicine.

    She is internationally recognized for her work on weight regulation and obesity throughout the adult lifespan, with a preventing unhealthy aging through weight management. She co-leads the NIH POWERS consortium, a discovery science initiative to identify determinants of weight regain following weight loss. She has also developed a novel remotely-delivered behavioral intervention that is highly effective for sustainable weight loss through dietary change, and has also demonstrated that her program can influence the reward system of the brain, resulting in increasing liking for healthy foods and reduced liking for unhealthy foods.

    In addition to her work in the U.S. she leads an international consortium of scientists in 9 countries (Brazil, China, Ghana, Guinea-Bissau, Kuwait, Finland, India, Italy, Singapore) dedicated to developing effective interventions for adult obesity worldwide. She has also worked for several years with co-investigators in Guinea-Bissau and Uganda on child undernutrition and cognitive effects of nutritional In the U.S., Dr Roberts has sat on national and international committees for dietary recommendations including the National Academies of Sciences Dietary Reference Intake panel, and has published over 300 research papers. She is the awardee of pre-eminent awards for national and international nutrition research including the 2009 E.V. McCollum award and 2024 E.V. McCollum International Nutrition Lectureship award of the American Society for Nutrition, and the USDAs 2016 W. O. Atwater Lecturer award.

Peter Solberg, MD

Chief Health Information Officer at Dartmouth Health

Keith Paulsen, PhD

Robert A. Pritzker Professor of Biomedical Engineering at Thayer School of Engineering

Moderator

Michael Whitfield, PhD

Chair of Biomedical Data Science at Dartmouth College