Gabrielle Hoyer
PhD Candidate | ML/CV for Medical Imaging | Foundation Models | Digital Twins
I build foundation-model and multimodal ML systems for healthcare. Current focus: musculoskeletal MRI, longitudinal digital twins, and clinical video pipelines for decision support.
Contact MeAbout Me
My introductionPhD Candidate (UCSF and UC Berkeley) building ML/CV systems for clinical data with an emphasis on foundation models, multimodal representation learning, and longitudinal decision modeling. I build end-to-end pipelines that span data curation, QA/QC, reproducible training and evaluation, and deployment-aware validation. My goal is to develop models that generalize across sites, protocols, and patient populations, and to produce outputs that remain useful for both patient-facing decisions and hospital operations. My current work centers on musculoskeletal imaging, and I apply the same systems approach to other clinical data types, including video and clinical text.
Areas of Focus
- Foundation models for MRI: robustness, biomarker fidelity, and evaluation tied to clinical endpoints
- Multimodal + longitudinal modeling (imaging + tabular + reports) for risk prediction and digital twins
- Data/ML systems for real-world translation: QA/QC, reproducibility, privacy, and workflow constraints
Stakeholder Interviews
Research
Strategic Leadership
Research Highlights
Flagship papers, preprints, and works in progress
Offline RL on a Longitudinal Knee Digital Twin
Clinically informed intervention design with offline RL (CQL/IQL) + off-policy evaluation on longitudinal biomarker trajectories.
MSK MRI Segmentation and Registration
Invited to teach quantitative MSK MRI workflows spanning segmentation, registration, and biomarker-ready analysis.
Qualification
My personal journeyBS in Biomedical Engineering, Neural Interfaces
University of UtahMS in Biomedical Engineering, Neural Interfaces
University of UtahPhD in Bioengineering, Machine Learning
UCSF-UC BerkeleyUndergraduate Research Assistant
Neural Information Lab, University of UtahUndergraduate Research Assistant
Mechanisms of Synaptic Functions Lab, University of UtahProduct Development Engineering Intern
Ortho Development, SLC, UTData Science Impact Fellow
Sorenson Impact Center, SLC, UTGraduate Researcher
Computer Vision in Anesthesiology Lab, University of UtahSoftware Development Intern
Compassion International, Colorado Springs, COTechnical Lead
A.I. Healthcare Venture, NSF National I-Corps Program, Winter CohortPhD Researcher
Musculoskeletal Quantitative Imaging Research Group, UCSF, CAVolunteer Lab Technician
Orthotics and Prosthetics Department, Shriners Hospital for Children - SLC, UTDirector of Research + Project Development
Project Embrace nonprofit - SLC, UTCollege of Engineering Ambassador
University of Utah - SLC, UTChief Development Officer
Project Embrace nonprofit - SLC, UTResearch Projects
What i'm working onDigital Twin Systems
& Precision Medicine
Digital Twin Systems & Precision Medicine
-
Developing personalized digital twin systems that integrate multi-modal clinical data (imaging, clinical demographics, biomechanical signals, and 3D anatomical representations) for precise musculoskeletal health management.
-
Applying advanced causal inference, predictive analytics, and generative AI to perform longitudinal analyses and individualized scenario simulations.
-
Creating interactive decision-support tools that allow clinicians and researchers to explore preventive care strategies, disease trajectories, and personalized treatment plans.
-
Establishing analytical methods and software platforms to enable real-time clinical translation and improve patient-specific clinical outcomes.
- GitHub Slide Deck Video Paper
Foundation Models &
Model Generalization
Representation Learning
& Efficient Data Utilization
Medical Data Systems:
Strategic Insights & Clinical Operations
Medical Data Systems:
Strategic Insights & Clinical Operations
-
Designing end-to-end systems to collect, clean, and analyze clinical data—including automated preprocessing of procedural video—for real-time quality assessment, privacy protection, and scalable downstream analysis.
-
Using performance and procedural metrics (e.g., video laryngoscopy outcomes) to uncover workflow inefficiencies, support training, and drive process refinement across diverse clinical environments.
-
Collaborating with clinicians, administrators, IT teams, and insurers to align AI tools with hospital workflows, operational goals, and implementation constraints.
- Demo GitHub GitHub
Previous Work
Projects + Research
Parkinsonism Biosignals
Analyzing cortico-subthalamic beta-band coherence and movement to study Parkinsonian motor states.
Poster
Let's Work Together
Join me in creating AI solutions to alleviate healthcare provider burnout, extend care to underserved communities, and enhance patient quality of life. Let's make a positive impact together.
Contact Me
Academic Works
Publications and AwardsDesign and Evaluation of Automatic System for Medical Dataset Development
Hoyer, G., Tong, M.W., Bhattacharjee, R. et al. Clinical Utility of Foundation Segmentation Models in Musculoskeletal MRI: Biomarker Fidelity and Predictive Outcomes. (2025). (under peer review).
Foundations of a Knee-Joint Digital Twin
Hoyer, G., Gao, K.T., Gassert, F.G. et al. Foundations of a knee joint digital twin from qMRI biomarkers for osteoarthritis and knee replacement. npj Digit. Med. 8, 118 (2025).
Fully automatic thigh muscle segmentation for the Osteoarthritis Initiative: a foundation model approach.
Ziegeler K.*, Hoyer, G.*, Liu, F., et al. Fully automatic thigh muscle segmentation for the Osteoarthritis Initiative: a foundation model approach. (2025). (under peer review).
Knee joint effusion-synovitis volume quantification on MRI.
Marth, A.A.*, Liu, F.*, Hoyer, G., et al. Knee joint effusion-synovitis volume quantification on MRI. (2025). (under peer review).
An Evaluation on Model Fairness in a Self-Supervised Learning Paradigm
Huang, H., Hoyer, G., Hess, M. et al. Demographic Bias Always Affect Model Fairness? An Evaluation with Musculoskeletal Imaging. (2025).
Validation of the Clinical Viability of ML Tools and Automated Analysis: A Case Study in Thigh Muscle Segmentation.
Liu, F., Hoyer, G., Ziegeler K. et al. Validation of a Fully Automized Volumetric Thigh Muscle Segmentation Pipeline Using Segment Anything. In Proceedings of the OARSI World Congress, Incheon, South Korea, (2025).
MskSAM - Foundation model for Musculoskeletal MRI Segmentation
Hoyer, G. & Tong, M.W. et al. Towards a Generalizable Foundation Model for Multi-Tissue Musculoskeletal MRI Segmentation. In Proceedings of the 32nd Annual Meeting of ISMRM, Session: The Future of AI in MRI: Emerging Technologies & Directions, Singapore, (2024), 8045. (Awarded Magna Cum Laude).
Quantitative MRI Interpretable 100D Feature Space
Hoyer, G., Gao, K.T., et al. Quantitative MRI Interpretable 100D Feature Space of Knee Osteoarthritis. In Proceedings of the 31st Annual Meeting of ISMRM, Session: ML/AI New Ideas, Toronto, Ontario, Canada, (2023), 993.
Comparison of Procedural Distancing in Airway Management
Hoyer, G., Runnels, S. et al. Comparison of Procedural Distancing of Primary and Advanced Intubation Techniques. In Proceedings of the Difficult Airway Society Annual Meeting, UK, (2020).
Collection and Strategic Use of Intubation Data
Hoyer, G., Runnels, S., Kuck, K. et al Automatic Video Laryngoscope Archiving System, First Pass Rates in Archived Intubations. In Proceedings of the Society for Technology in Anesthesia Annual Meeting, Austin, TX, (2020).
Investigating Beta-Coherence in Parkinsonian Rats: Self-Directed Motion Effects
Polar C.* & Hoyer, G.*, Dorval A., Cortico-Subthalamic Beta-Coherence Suppression During High-Intensity, Self-Directed Motion in a Parkinsonian Rat Model. In Proceedings of the 9th International IEEE/EMBS Conference on Neural Engineering (NER), IEEE Engineering in Medicine and Biology Society, San Francisco, CA, (2019), 491.
Towards the Provision of Sustainable Medical Materials to Low and Middle-Income Countries
Sudabattula, M. & Hoyer, G. Towards the Provision of Sustainable Medical Materials to Low and Middle-Income Countries: A Non-Profit Case Study. In Proceedings of the 2017 International Health Conference, Session: Health Inequalities and Big Data, St. Hugh's College, Oxford, UK, (2017), 178.
Google PhD Fellowship Nominee
2025Nominated for the Google PhD Fellowship, a highly competitive program recognizing outstanding doctoral researchers. Selected by UCSF as a nominee for my work in AI for health and human-centered tools. Proposal: JointAI, a knee digital twin that uses MRI biomarkers to track disease progression, estimate risk, and simulate interventions.
Google PhD Fellowship Nomination - Proposal: JointAI
-
Digital twin of the knee built from clinical MRI and quantitative MRI biomarkers.
-
Self-supervised representation learning for imaging features; longitudinal models for risk stratification.
-
Causal inference and reinforcement learning for scenario testing and decision policies.
-
Interactive 3D visualization platform for researchers and clinicians.
- Research Proposal PDF
Fletcher Jones Fellowship Nominee
2025Nominated for a competitive research fellowship recognizing outstanding dissertation research and potential for significant contributions to health sciences.
UCSF Discovery Fellowship
2023 – PresentDiscovery Fellows are chosen for their leadership potential, excellence in research, community-mindedness, and communications skills. The role entails advocating for UCSF's foundational and biomedical science programs while championing innovative research within healthcare.
NIH T32 BioE Training Grant
2021 – 2022NSF National I-Corps Program Grant
2021The National Science Foundation's Innovation Corps (I-Corps™) program offers experiential education to equip researchers with valuable insights into entrepreneurship, business startup, and industry demands and challenges. I-Corps serves as a catalyst for turning inventions into real-world impact, combining scientific exploration with industrial discovery in a rigorous, relevant, and evidence-based environment. This program empowers researchers to accelerate the transition of lab concepts to market-ready solutions, contributing to a national innovation network that transforms scientific research into societal benefits.
NASA Space Grant Fellowship
2019 – 2021Initiated in 1989, the National Space Grant College and Fellowship Project, or Space Grant, is a national network of colleges and universities aimed at expanding opportunities for Americans to engage in NASA’s aeronautics and space projects. This includes supporting science and engineering education, research, and public outreach efforts to enhance understanding and participation in space exploration and related fields.
Chevron Scholarship for Women in Stem
2018 – 2019The Chevron Scholarship for Women in Science, Technology, Engineering, and Math (STEM) fields supports self-identified women majoring in STEM. Created for the Women's Resource Center, this scholarship aims to encourage and assist women pursuing education and careers in science, technology, engineering, or math.
Bioscience Undergraduate Research Scholar
2015 – 2017Awarded by the Department of Biology at the University of Utah, this recognition supports and acknowledges outstanding undergraduate students engaged in bioscience research. The program aims to foster a deeper understanding and passion for scientific inquiry among bioscience students through active participation in research projects and initiatives.
NSF Research Experiences for Undergraduates Grant
2015 – 2016Associated with the Neural Information Lab and Dr. Chuck Alan Dorval, the NSF Research Experiences for Undergraduates (REU) program encourages active research participation by undergraduate students in areas funded by the National Science Foundation. This initiative aims to provide undergraduates with hands-on research experience, fostering a deeper engagement and understanding in their field of study.
ACCESS Program for Women in STEM Scholar
2014 – 2015A scholarship award and first-year experience designed specifically for incoming fresh-women with a keen interest in the sciences, including math, biology, chemistry, or physics. The ACCESS Program supports and encourages women embarking on their journey in STEM fields, aiming to provide them with foundational support, mentorship, and engagement in their chosen disciplines.