The rise of the virtual twin heart is rapid. For example, researchers want to practice operations with this. ‘You can try a new procedure a thousand times on a computer heart.’
Elleke Bal22 February 2024, 03:00
A patient comes to the doctor. He is short of breath, has chest pain and tires easily. It’s very wrong, the cardiologist tells the man that he suffers from heart failure. He even needs a new heart valve. His heart is imaged with an ultrasound.
When the patient is back in the consultation room to discuss the operation, the cardiologist suddenly turns the computer screen towards him. The patient blinks briefly.
Because there, on that screen, he sees his own heart beating. Based on the ultrasound images, the computer has calculated how his specific heart functions and responds. You could call it his twin heart. The ventricles and atria contract as only happens in the heart. In that virtual copy, the doctor tried out types and sizes of valves to see how his heart would respond in the future. The valve that produces the best results is placed in his heart a week later.
Predicting disease progression
For the time being, things are not going that way in the consulting room. But that won’t last long, because research into virtual heart copies is moving fast.
It is not surprising that researchers and companies in the medical world are pushing ahead. The models cannot only be used to select the perfectly fitting heart valve. Also consider testing medication, or predicting disease progression by accelerating the development of such copies.
In the meantime, the rise of these computer models also raises questions. Because: how accurate are the copies, and how will doctors use them? Nobody wants a model that makes an incorrect prediction on which a cardiologist bases a judgment.
Pulsating hearts
Belgian Mathias Peirlinck has been working on modeling hearts as a biomechanical engineer for eleven years, and he is still amazed every day at how ingenious it is. Consider the way the heart pumps blood from the ventricles. “The heart rotates a little and contracts at the same time, like wringing out a wet towel. That rotating movement is simply very efficient.” Take two water bottles, he says, and empty them, then the bottle you rotate will also empty faster.
Peirlinck has had his own lab in Delft, Netherlands, for three years. Not a room full of bells and whistles, gigantic machines or research equipment. He usually doesn’t need it, because his main tool is the computer. His colleagues work in bare white office spaces at the mechanical engineering faculty of TU Delft. Every now and then Peirlinck shows pictures of pulsating hearts on his computer screen.
Anyone who reads about the workings of the heart in any biology book might just think that we know exactly how it works. Anyone who talks to Peirlinck for a while will doubt that. How cells in heart tissue connect to each other and communicate is still far from understood, he says. Just as we do not know the mechanical strength of the tissue well. Research into how electrical stimuli stimulate the heart rhythm via a network of cells is also still in development.
As an engineer, Peirlinck ultimately wants to understand how the heart becomes diseased, and how heart tissue evolves over time. For example, hearts start to fail if blood pressure is high for a long time, he explains. In response to that blood pressure, the wall becomes thicker, because the heart wants to be able to handle that force better. “That is the biomechanical response to overload. But eventually the pumping force decreases, and the heart finds it increasingly difficult to do its work.”
It is not clear how this wall is induced to stiffen at cell level. Peirlinck therefore wants to understand which mechanical forces are at work. To do this, he sometimes uses studies on healthy and diseased heart tissue that have been previously done by other scientists. Pieces of heart tissue were pulled in different directions and the forces were measured. “We then want to understand the relationship between the forces and the deformation of the tissue: that is mechanics.”
Computing power
Yet he touches very little heart tissue and his work is largely computer-controlled. There is no other way, he says, because most forces cannot be seen or measured, only calculated. “To calculate all the forces in a living organ, with countless muscle cells that all behave and interact differently, you need the computing power of a computer.” Peirlinck summarizes the functioning of thousands of small pieces of tissue in mathematical equations that he then calculates.
He does this on the basis of the biological, electrophysiological and mechanical knowledge that he experimentally collects about that tissue. Monk work.
Before a virtual copy can completely map a human heart from cell to organ level, a lot of calculations still have to be done. But for many applications it is not necessary to fully understand the heart, says Peirlinck. For example, to test a new heart valve, a somewhat simpler model will also suffice. Computers then use information from an ultrasound, CT scan or MRI scan to create a model. “We make this virtually correct based on our experimental knowledge and principles of physics.”
Personalized model
These digital copies cannot be called 3D models, says Peirlinck, because it goes a step further. It is more like 4D, because a fourth dimension has been added. The model is dynamic and can, for example, be flushed to study the deformation of the heart over time.
“Precisely because of this, the model can predict how the heart will respond to treatment or a certain type of heart valve. Every heart is different, so it adds a lot if a doctor can, for example, try out valves in a personalized model. One valve is too narrow, the other places too much tension on the tissue. There is a real difference.”
But his team is not only working on the ultimate digital clone, with which they can answer fundamental questions about the heart, but also on models that can be used more quickly in practice. For example, he shows a beating heart on the computer with a winding pacemaker wire inside. In some patients, the wire breaks after a while, causing the device to no longer work. The team can use the computer to test what the cause is. You can then virtually test with materials of different thickness or stiffness to see how the wire could last longer.
Other models can, for example, be used by doctors to test operations. They can then, as it were, run ‘trial laps’ on the computer before they start the real operation.
Role of doctor
Such a model should never take over the doctor’s decision-making role, says Peirlinck. He emphasizes that the models should be seen as additional data, allowing a doctor to make a better decision. “And all models must first be validated by proving their effectiveness in scientific research.”
Even then, such models are not simply allowed into the consultation room, he says. There are extensive European and international safety procedures and guidelines for this.
On the other hand, says Peirlinck, the models also reduce risks. “No one dies when the computer heart fails. You can try a new procedure a thousand times on a computer heart, or test ten different heart valves.” The side effects of a medicine can also be tested more safely via a computer heart than via a living patient.
“I especially hope that we can work in an increasingly personalized way,” says Peirlinck. “Doctors now have to make treatment decisions based on studies conducted on a hundred people. And those results are then generalized to the entire population. There are also risks involved. By using an individual model of each heart, you can make a decision that fits that unique heart much better.”
Peirlinck continues to puzzle to make the most accurate copies of all those unique hearts available. And like many scientists before him, he must first try to understand how incredibly clever nature is. “We have not yet developed a single pump that works as efficiently and for a long time as the human heart itself.”
The effect of a pacemaker
Researchers in Maastricht are also working on building a digital heart. Their model recently showed that it successfully predicts whether treatment with a pacemaker is effective.
The study involved 45 patients with heart failure. A digital copy of the heart was made for each of them, Joost Lumens explains. He is a professor of biomedical technology and led the study. “Our model is somewhat coarser than the model being built in Delft. An advantage of this is: it can quickly create a personalized model of the pump function, based on ultrasound images and medical data.”
The 45 patients had heart failure with a conduction disorder in the electrical signal of the heart. A pacemaker could solve that, but it does not work in all cases, says Lumens. “Many patients already have a weak heart muscle. And then you can stimulate, but that muscle does not respond well.” The digital model can calculate the properties of that muscle, how strong or weak it is.
The 45 patients received the pacemaker in accordance with the guidelines. The outcomes of the treatment were compared with the predictions of the model. And they turned out to be very similar.
In the future, the optimal setting of the pacemaker could possibly also be predicted. Although it is not yet permitted to adjust treatment based on a prediction of a digital twin heart. Lumens: “Clinical guidelines need to change for this.”
The soft robot heart
Heart failure can be treated with medication and surgery, but a heart transplant is sometimes the only cure for the most serious cases. However, there is still a major shortage of donor hearts. And that is why researchers are busy building artificial alternatives to the human heart, such as the temporary support heart.
Last year, the Holland Hybrid Heart consortium received 10 million euros from the National Science Agenda to build another, temporary alternative for people on the transplant list: a robotic heart with soft walls. “If you want to build a new heart pump, you cannot use hard materials, as that does not combine well with the blood flow,” says Mathias Peirlinck, who also uses his knowledge about how the heart works for this research. “Digital clones are also interesting for this research: we can virtually test prototypes and calculate how they would function in a real body.”
The research is led by cardiothoracic surgeon Jolanda Kluin from Erasmus MC, and she has set the goal: to introduce a soft robotic heart within ten years as a useful alternative to heart transplants with donor hearts.
Also read
2024-02-22 02:00:31
#dies #computer #heart #fails #Belgian #digital #twins #heart #Delft
