From Monitoring to Prediction: How Cardiovascular Digital Twins Could Transform Arrhythmia Care

Interview With Amanda Randles, PhD

© 2026 HMP Global. All Rights Reserved.

Any views and opinions expressed are those of the author(s) and/or participants and do not necessarily reflect the views, policy, or position of EP Lab Digest or HMP Global, their employees, and affiliates. 

EP LAB DIGEST. 2026;26(3).

Interview by Jodie Elrod

In this EP Lab Digest interview, Amanda Randles, PhD, director of Duke University’s Center for Computational and Digital Health Innovation and head of the Randles Lab, explores how cardiovascular “digital twins” could move arrhythmia care beyond reactive monitoring toward proactive, personalized management. By integrating longitudinal rhythm data with patient-specific anatomy, hemodynamics, and physiology, digital twins offer a systems-level view of arrhythmia risk—revealing hidden contributors, anticipating vulnerability before symptoms arise, and enabling clinicians to virtually test treatment strategies such as ablation, pacing, and drug therapy tailored to the individual patient.

Much of digital health in cardiology focuses on monitoring and alerts. How does your concept of a cardiovascular “digital twin” fundamentally change how we might understand and manage arrhythmias?

We have seen tremendous advances in arrhythmia care through continuous and remote monitoring. Wearables and implantable devices have fundamentally changed our ability to capture rhythm disturbances outside the clinic, and they have clearly shown that longitudinal, real-world monitoring can meaningfully impact outcomes. Simply being able to follow arrhythmias as they evolve over days to months rather than at isolated clinic visits has already reshaped how we diagnose disease, stratify risk, and manage patients.

What this progress makes clear is that continuous data matter. But monitoring alone keeps care largely reactive. We detect events once they happen and respond after the fact. A cardiovascular digital twin builds directly on the success of monitoring and represents the next step in using longitudinal data to model an individual patient’s underlying physiology and anticipate what comes next. Rather than alerting when a rhythm has already crossed a threshold, a digital twin aims to identify early physiologic changes that increase vulnerability to arrhythmias before they become clinically apparent. By integrating rhythm data with anatomy, hemodynamics, activity, and other patient-specific factors, the focus shifts from event detection to risk anticipation.

Digital twins also give clinicians a new kind of decision support tool. Instead of relying solely on population-level trial averages or trial and error adjustments, physicians could explore how different medications, pacing strategies, ablation targets, or lifestyle changes are likely to affect this patient’s rhythms ahead of time. In this way, digital twins move arrhythmia care toward proactive, personalized management, where decisions are guided by predicted benefit rather than retrospective response.

What insights can patient-specific cardiovascular models provide about the structural or hemodynamic contributors to arrhythmias that may not be apparent on standard imaging or electrocardiograms (ECGs)?

Standard imaging and ECGs are incredibly valuable, but they only give us snapshots. A patient-specific cardiovascular model lets us connect those snapshots into a coherent physiologic picture and uncover contributors to arrhythmias that are otherwise hidden.

Structurally, digital twins can capture how subtle geometric features such as chamber dilation, regional wall stress, or fibrotic burden interact with an individual’s anatomy in ways that are difficult to infer from images alone. Two patients may look similar on magnetic resonance imaging or echocardiography, but experience very different electrical behavior because of how those structures alter local loading, strain, or conduction pathways.

From a hemodynamic perspective, models let us quantify forces and pressures that are not directly measurable, such as spatially varying filling pressures, wall shear stress, or changes in preload and afterload over time. These hemodynamic patterns can create a permissive environment for arrhythmias by altering myocardial stretch, oxygen demand, or autonomic balance long before overt rhythm changes appear on an ECG.

Importantly, patient-specific models also allow us to study how these structural and hemodynamic factors evolve together over time. Rather than viewing arrhythmias as isolated electrical events, we can begin to understand them as emergent behaviors of a coupled mechanical, hemodynamic, and electrical system. That systems-level view is what enables earlier risk identification and more targeted, personalized intervention strategies.

Do you see digital twins helping clinicians identify patients at risk for atrial fibrillation or ventricular arrhythmias before clinical symptoms or sustained episodes occur? Could digital twins also eventually help predict how an individual patient’s heart might respond to ablation, antiarrhythmic drugs, or pacing strategies?

Great question, that’s exactly where I think the promise of digital twins lies. They give us a way to move upstream from detecting arrhythmias to identifying vulnerability. By combining longitudinal rhythm data with patient-specific anatomy, hemodynamics, and physiology, digital twins can highlight early changes that increase arrhythmia risk well before sustained episodes or symptoms occur.

They also allow us to move away from population-derived thresholds and instead understand how a particular patient has shifted from their own baseline. A measure that might not look remarkable on its own could be meaningful if it represents a large deviation for that individual. In the same way, a change that seems small at the level of a single monitored signal can have downstream effects on other physiologic factors. Those effects can be synergistic and amplified, and that is something a digital twin is designed to capture but a monitor alone may miss.

I also think digital twins open the door to much more personalized treatment planning. Because these models are built around an individual patient, they create a framework for exploring how that patient’s heart is likely to respond to different interventions before anything is done clinically. Over time, this could allow clinicians to test ablation strategies, antiarrhythmic drugs, or pacing approaches virtually and get patient-specific insight into what is most likely to work. The goal is not to replace clinical judgment, but to reduce trial and error and help clinicians make more informed, personalized decisions earlier in the course of disease.