Duke Digital Twin Initiative
Healthcare is shifting from reactive treatment to proactive, predictive care. Digital twins—dynamic, data-driven representations of individuals that integrate real-time data, computation, and modeling—enable earlier detection, personalized intervention, and continuous monitoring.
However, the underlying data and tools remain fragmented. Imaging, wearable data, clinical records, behavioral signals, and computational models are developed and analyzed in isolation, limiting their ability to capture the full trajectory of human health.
The Duke Center for Computational and Digital Health Innovation (CCDHI) is actively bringing these pieces together. By integrating expertise across medicine, engineering, high performance computing, artificial intelligence, imaging, and digital health, CCDHI is establishing a cohesive and scalable foundation for digital twin science.
A defining strength of this effort is Duke’s leadership in advanced imaging. Quantitative MRI, functional imaging, and large-scale image analysis provide high-resolution insight into structure and physiology—forming the basis for accurate, mechanistic digital twins.
The Initiative
The Duke Digital Twin Initiative, led by CCDHI, is building an integrated platform that brings together multimodal data, physics-based modeling, and AI to enable human digital twins at scale. This platform supports earlier disease detection, prediction of outcomes, and more precise, personalized care across domains.
Rather than focusing on a single condition, the initiative is advancing high-impact applications across:
- Cardiovascular health and heart failure. Digital twins integrate imaging, wearable data, and computational modeling to track hemodynamic changes over time—enabling earlier detection of decompensation and more informed intervention.
- Neurological disease and cognitive decline. By combining imaging, gait, eye tracking, and behavioral data, digital twins improve the differentiation of disorders such as Alzheimer’s, Parkinson’s, and NPH while allowing longitudinal monitoring.
- Sports and human performance. Digital twins support athlete monitoring through wearable data, motion analysis, and physiological modeling—advancing injury prevention, recovery, and performance optimization.
Across these domains, the approach is consistent: shifting from static snapshots to continuous, personalized models that evolve over time.
The Backbone: PRISM
At the core of this initiative is PRISM (Platform for Research Integration of Secure Multimodal Modeling), developed within CCDHI as a secure, FAIR, AI-enabled platform for digital twin research and deployment. PRISM integrates electronic health records, imaging, wearable data, motion and gait analysis, behavioral signals, and computational models into a unified environment for analysis and simulation.
By providing a reusable infrastructure, PRISM enables rapid development, validation, and deployment of digital twins across applications—eliminating the need for fragmented, disease-specific pipelines.
What Makes This Approach Different
- Imaging-driven foundations. Advanced imaging provides quantitative insight into anatomy and physiology. Combined with segmentation and modeling, these data enable reconstruction of patient-specific structure and function.
- Longitudinal wearable integration. Wearables capture continuous signals such as activity, heart rate, and sleep—allowing digital twins to evolve with the patient over time.
- Behavioral and functional phenotyping. Eye tracking, gait analysis, and video-based assessment capture subtle changes in cognition and movement beyond traditional clinical metrics.
- Physics-based modeling. Mechanistic models derived from imaging enable simulation of physiological processes and prediction of intervention outcomes.
- Unified AI-enabled platform. PRISM integrates these data streams into a single environment, enabling multimodal analysis and discovery of composite biomarkers.
Why Duke Will Lead
Through CCDHI, Duke combines the capabilities required to lead digital twin development and deployment:
- Integrated ecosystem: Seamless collaboration across engineering, medicine, nursing, and data science.
- Imaging leadership: Advanced MRI, quantitative imaging, and analysis expertise.
- Computational strength: Leadership in high performance computing, AI, and modeling.
- Clinical scale: Access to diverse patient populations and longitudinal data.
- Proven innovation: Established track record in digital twins, wearables, and translational research.
These strengths enable Duke not only to develop digital twins, but to deploy them in real-world settings.
Transformative Impact
The Duke Digital Twin Initiative establishes a unified framework that integrates imaging, physiology, behavior, and modeling into a single system. This enables identification of multimodal biomarkers, earlier detection of disease, and more precise, personalized care.
By building a reusable, scalable platform through CCDHI, Duke is accelerating digital twin development across domains—from cardiovascular disease and neuroscience to human performance. This initiative positions Duke as a national leader in digital twin research and deployment and creates a durable foundation for innovation across medicine and health.
Focused Pilots: Turning the Vision into Practice
The Duke Digital Twin Initiative is launching a series of focused pilots designed to demonstrate how a shared digital twin infrastructure can address specific, high-impact challenges in human health. Each pilot brings together clinical expertise, multimodal sensing, imaging, computational modeling, and AI, while contributing capabilities that can be extended across future applications.
First Pilot: A Digital Twin for Heart Failure
Our first flagship pilot focuses on heart failure, with the goal of moving from episodic clinical measurements toward continuous, longitudinal understanding of cardiovascular health. The project integrates cardiovascular imaging, wearable and implantable sensors, physiological measurements, and high-fidelity computational models to create a digital twin that evolves with the patient over time.
The long-term vision is to identify physiological deterioration before symptoms or hospitalization, giving clinicians an opportunity to intervene earlier and enabling a shift from reactive treatment toward proactive, personalized cardiovascular care.
Additional focused pilots spanning other areas of health and human performance are in development. More information will be announced as these programs launch.
Join the Duke Digital Twin Initiative
Our ambition is to move digital twins from promising research to technologies that can improve health at scale. Over the next decade, our goal is to help digital twin technologies developed through this initiative reach one million people.
Achieving that goal will require more than any single laboratory, discipline, or institution. We are building a community of researchers, clinicians, technology developers, industry partners, and supporters who want to help create the next generation of personalized, predictive health technologies.
Collaborate with us.
We are seeking researchers and clinicians interested in contributing expertise, data, technologies, or new clinical applications. If you are developing sensing technology, imaging methods, computational models, AI, software, or other capabilities that could strengthen a digital twin, we want to explore how they could become part of the initiative.
Partner with us.
We welcome industry partners interested in developing, validating, translating, or scaling digital twin technologies. Partnerships can range from integrating new hardware and software into our focused pilots to jointly developing capabilities that can ultimately reach patients at scale.
Support the initiative.
Philanthropic support provides the flexible resources needed to launch ambitious pilots, connect teams across disciplines, build shared infrastructure, generate the evidence required for larger clinical studies, and move promising ideas toward real-world impact.
Help us reach one million people.
