Researchers have created more than 3,800 anatomically accurate digital replicas of human hearts, revealing how age, sex and obesity influence the heart’s electrical behaviour. The technology could ultimately contribute to more personalised treatment of cardiovascular disease.

What if scientists could study thousands of human hearts without performing a single invasive procedure?

Researchers from Imperial College London, King’s College London and The Alan Turing Institute have taken an important step in that direction by creating more than 3,800 cardiac digital twins.

Each is a sophisticated computer model based on data from a real person, designed to reproduce important aspects of that individual’s heart.

It is the first time cardiac digital twins have been created and studied on this scale.

What is a digital twin?

The idea originated in engineering.

A digital twin is a computer model representing a real physical object or system. Engineers can use such models to investigate how machines behave without having to perform every experiment on the physical machine itself.

Medicine is now adopting the same principle.

A cardiac digital twin can reproduce characteristics of an individual patient’s heart, allowing researchers to investigate processes that are difficult to measure directly.

The Imperial-led team created its models using real patient data and electrocardiogram (ECG) measurements from the UK Biobank and a group of patients with heart disease.

Building thousands of virtual hearts

Creating an accurate computational heart traditionally requires considerable time and specialist work.

Recent advances in machine learning and artificial intelligence have changed that.

The researchers automated parts of the process, making it possible to construct digital hearts much faster and therefore study thousands of them simultaneously.

That scale is important.

Rather than examining only one virtual patient, scientists can now investigate differences across an entire population.

Age changes the electrical heart

One of the clearest findings concerned ageing.

The digital twins showed that age influences the electrical properties of the heart.

That is important because the heartbeat is controlled by precisely coordinated electrical signals. Disturbances in those signals can contribute to abnormal heart rhythms and other cardiovascular problems.

The models therefore provide researchers with another way to investigate why the risk of heart disease changes as people grow older.

Obesity leaves its mark too

Age was not the only factor.

The researchers also found that obesity is associated with changes in the heart’s electrical properties.

This could help explain some of the relationship between obesity and increased cardiovascular risk.

Instead of merely observing that obesity correlates with heart disease, digital twins offer scientists a way of investigating what may actually be changing inside the heart.

Men and women: an unexpected finding

The researchers also investigated a familiar phenomenon in cardiology.

Doctors have long observed differences between the ECG readings of men and women.

One possible explanation would be that male and female hearts conduct electrical signals differently.

But the digital twins pointed towards a simpler explanation.

The study found that differences in ECG measurements between men and women were primarily explained by differences in heart size, rather than fundamental differences in the way their hearts conduct electrical signals.

That distinction could matter clinically.

If differences in ECG readings are partly caused by anatomy rather than electrical biology itself, doctors may eventually be able to refine how particular measurements are interpreted and how treatments are configured.

From population research to personalised medicine

Professor Steven Niederer, Chair in Biomedical Engineering at Imperial’s National Heart and Lung Institute and senior author of the research, argues that the importance of digital twins extends beyond diagnosis.

By creating virtual hearts representing people across a population, researchers can investigate which groups may be particularly vulnerable to heart disease and how factors such as lifestyle influence cardiac function.

The approach could ultimately help doctors personalise treatment.

One possibility mentioned by Imperial is adjusting the settings of cardiac devices more precisely for different patients.

Another is using the models to identify potential drug targets for particular groups.

A virtual laboratory for the human heart

Perhaps the most powerful aspect of the technology is that a digital twin can function as a kind of virtual laboratory.

Some properties of a living human heart are extremely difficult to measure directly.

A computational replica allows researchers to investigate those properties without exposing a patient to an invasive experiment.

In the future, increasingly sophisticated digital twins could potentially help scientists investigate how a disease develops or how different patients might respond to particular treatments.

That does not mean a computer model can already predict everything that will happen to an individual patient. Digital twins remain models and must be validated carefully against real clinical evidence.

But their ability to combine real patient data with sophisticated simulation is opening a new way of studying medicine.

From one heart to thousands

The significance of the Imperial-led study lies partly in its scale.

A single digital heart can tell researchers something about one patient.

Thousands of digital hearts can begin to reveal patterns across populations.

And those patterns can then be related to age, sex, obesity and other characteristics.

Professor Pablo Lamata of King’s College London, one of the authors, said that developing the technology at scale opens the possibility of using digital twins in large population studies and ultimately developing better prevention strategies and more personalised treatments.

The heart enters the digital age

For centuries, understanding the human heart depended largely on observing the physical organ itself.

Then came the stethoscope, the electrocardiogram, ultrasound, CT and MRI.

Digital twins represent another step.

Researchers are no longer limited to observing what a heart is doing. Increasingly, they can construct a computational version of it and ask what might happen under different circumstances.

More than 3,800 virtual hearts have now helped scientists reveal how age, obesity and anatomical differences influence cardiac behaviour.

The next challenge is turning those digital insights into better decisions for real human hearts.

Source: Based on “Thousands of cardiac ‘digital twins’ offer new insights into the heart”, published by Imperial College London on 20 May 2025. The research involved Imperial’s National Heart and Lung Institute, King’s College London and The Alan Turing Institute and was published in Nature Cardiovascular Research.