Building A Soft Robotic Heart That Mimics HFpEF Progression

By Pooja Toshniwal PahariaReviewed by Lauren HardakerJun 10 2026

A soft robotic heart model has recreated the pressure changes, filling abnormalities, and chamber interactions seen in HFpEF, giving researchers a powerful new way to study one of cardiology’s most complex diseases and evaluate future device-based treatments.

Study: Compliance modulation of a soft robotic atrioventricular model of heart failure with preserved ejection fraction. Image credit: gulfaraz783/Shutterstock.com

In a recent study published in Nature Communications, researchers developed a soft robotic atrioventricular model of heart failure with preserved ejection fraction (HFpEF). The model outlined key hallmarks of HFpEF and could help explain the biological mechanisms driving disease progression, thereby supporting the development of more targeted therapies and device-based interventions.

Researchers Build Soft Robotic HFpEF Heart Model

HFpEF is a complicated heart disease associated with multiple medical conditions, such as high blood pressure, irregular heart rhythms, kidney problems, obesity, and diabetes. It does not affect each individual in the same way. The heterogeneous nature and varied presentations of the disease make it difficult to design medical devices to improve cardiac pumping capacity.

Scientists are trying to fully understand how HFpEF begins and progresses. Existing animal models are expensive, time-consuming, and often require special diets or surgical procedures to recreate HFpEF. While laboratory-based simulators can better control blood flow and pressure changes, they have not fully replicated the heart's intricate mechanics.

Artificial Muscle Fibers Recreate Left Heart Function

In the present study, researchers developed an advanced laboratory model of the left side of the heart to study HFpEF progression.

The model comprised artificial muscle fibers to simulate the heart’s main pumping chamber (the left ventricle, LV), the left atrium (LA), and the circulatory components through which blood flows in the heart. The team prepared these artificial muscle fibers using hydraulic forces that could be programmed to replicate the functioning of a real human heart.

The artificial muscle fibers allowed precise control of the heart's filling phase (diastole). Using a feedback system and user-defined targets, the simulator continuously monitored the pressure inside the ventricle and adjusted the stiffness and relaxation states of these heart muscles. These functionalities allowed the researchers to study key features of HFpEF. The model reproduced changes that occur with increasing heart stiffness, slower relaxation between beats, and increased pressure in the chambers during filling.

The researchers initially tested this control system in individual artificial muscle fibers before applying it to the entire ventricle. They then added an artificial LA to study how the two chambers interact during heart filling. The model also included ultrasound imaging capabilities, allowing the team to visualize blood flow across the mitral valve in a way similar to clinical echocardiograms.

After replicating HFpEF changes in the muscle fibers, the team investigated whether a HeartMate 3 left ventricular assist device (LVAD), a form of mechanical circulatory support, could affect HFpEF. The simulator continuously adapted to changing pressure conditions and flow patterns. This enabled the team to assess how the system behaved under different cardiac loading conditions in a controlled laboratory setting.

Simulator Reproduces Hallmarks Of Early HFpEF

The model could closely mimic many changes that occur in HFpEF. The model accurately demonstrated biomechanical changes in the heart and alterations in blood flow. When the researchers simulated one of the earliest signs of HFpEF, i.e., the heart's reduced ability to relax between beats, the model showed that blood entered the heart more slowly and less efficiently. As a result, blood flow from the LA to the LV was delayed, reducing peak filling rates. The delay also increased LV pressure during diastole. This is similar to what doctors often observe in patients with HFpEF.

As the researchers increased ventricular stiffness and reduced the amount of blood it could hold during diastolic filling, the simulator reproduced features of more advanced HFpEF. The team observed worsening features of the condition, including increased filling pressure, decreased blood flow, and reduced ventricular volumes. The artificial heart muscle automatically adjusted its response to changing pressure targets, much like a real heart does as the disease progresses.

By adding an artificial LA, the team was also able to recreate the interactions between the heart's upper and lower chambers. The patterns of blood flow and pressure differences resembled clinical ultrasound findings in patients with HFpEF. Importantly, researchers could reproduce these findings at both 30 and 60 beats per minute (bpm).

The researchers further tested a HeartMate 3 LVAD as a form of mechanical heart support for HFpEF-like conditions. The simulator showed that increasing pump speed could relieve pressure and congestion in the LA. It also revealed changes associated with disease complications. For instance, when researchers increased the pump speed to excessively high levels, the LA collapsed.

Proof-Of-Concept Platform Advances HFpEF Research Tools

This proof-of-concept study shows that the heart simulator can more realistically recreate key changes seen in HFpEF than many existing laboratory models. The model reproduced interactions between different chambers of the heart and automatically adjusted to differences in cardiac stiffness and motion.

These model capabilities could help scientists study disease pathogenesis and test new therapeutic strategies, including mechanical circulatory support devices. However, the authors note that further development, validation, and operation at more physiologically relevant heart rates are needed before the platform can support clinical applications.

Researchers could improve the platform by more accurately simulating the entire heartbeat. They could also incorporate data from individual patients, other heart conditions, and advanced imaging and sensing technologies to support personalized therapies.

Download your PDF copy now!

Journal Reference

Davies, J., Sharma, B., Ji, A. et al. (2026). Compliance modulation of a soft robotic atrioventricular model of heart failure with preserved ejection fraction. Nature Communications, DOI: 10.1038/s41467-026-73791-w. https://www.nature.com/articles/s41467-026-73791-w