Medical digital twins are computerised versions of patients, enabling different treatments to be tested and the most effective one identified, without any risk to the patient's health. And in the future, we will each have our own, according to researcher Mikael Benson.
A common problem in healthcare is that it is often impossible to predict whether a particular patient should be treated with medicine A, B, or C. The only way to find out is to try the alternatives one by one, until something proves effective. Sometimes there are many options, and the search for a treatment that works can therefore take a long time, resulting in high costs, considerable suffering, and poorer health. In the worst case, waiting for an effective treatment can mean the patient's life can no longer be saved.
But what if the process of finding the right treatment could instead be simulated on a computer? That is the idea behind the digital twins being developed by Mikael Benson 's research group at the Department of Clinical Science, Intervention and Technology at Karolinska Institutet. Digital twins can be described as virtual test subjects, designed to mimic the individual patient. If the digital twin responds well to a particular treatment in the simulation, the principle is that the real patient is expected to benefit from the same treatment.
In the computer, treatments do not have to be evaluated one at a time; instead, they can be tested in parallel by a thousand cloned digital twins. I this way, healthcare providers receive immediate feedback on what is most likely to help a specific patient.
"Most diseases are extremely complex, which is why the same treatment has different effects on different people. Even the most trivial hay fever is caused by thousands of genes changing activity in billions of cells. These changes differ between patients with the same diagnosis, and so treatment outcomes also vary," explains Mikael Benson.
He estimates that roughly half of all prescribed treatments are ineffective.
Focus on cancer treatment
The digital twins developed by his group have, in recent years, been tested on mice with rheumatoid arthritis and on patients with Crohn's disease. In both cases, the results from the computer simulations closely matched real-world outcomes. The researchers also recently published a study in the scientific journal Cancer Research, where the methods were refined for the early diagnosis of prostate cancer.
Mikael Benson is now preparing a study to investigate how digital twins can help prevent cancer in patients with the bowel disease ulcerative colitis.
"Serious diseases with costly treatments, such as cancer and inflammatory bowel disease, could become important areas of use in routine healthcare. Our main research focus going forward is early diagnosis and treatment of cancer," he says.
For medical simulations with digital twins to work, large amounts of data about the patient's biology are required. The twin model is fed with information about exactly what activity is taking place in the patient's sick and healthy cells, as well as the patient's genetics, symptoms, and results from routine clinical examinations such as X-rays. Thousands of cells from the patient are examined individually using single-cell analysis to provide sufficient detail.
The collected data is then analysed. Using machine learning, patterns characteristic of how the disease manifests in the patient are identified.
To then proceed and test virtual treatments on the digital twins, information is also needed about how different medicines act at the molecular level. Such data exists for many pharmaceutical substances, but by no means all.
Digital twins are part of a broader trend in which computer simulations have, over recent decades, become an important tool in many fields, not least in various types of development work (see fact box).
Important not to uncritically adopt
In healthcare, digital twins can be designed in various ways for different purposes. While Mikael Benson's twins are a mathematical model, other researchers are developing digital twins that also have a visible exterior resembling the patient. These could be used as a basis for discussion
between doctor and patient. Different options become clear and concrete, for example, when you can see your digital twin on the screen losing a few kilos.
However, it is also important that healthcare does not uncritically adopt digital twins without reflecting on their use, Mikael Benson points out. The increased knowledge they can bring is not solely positive but can also lead to stress.
"Questions about lifestyle factors, such as smoking and exercise, will arise when doctors and patients look at different treatment options. The importance of the choices a patient makes becomes clear. This may lead to psychological pressure and even feelings of guilt. We need to be aware of that."
But overall, he sees opportunities.
What do you think about the long-term development of digital twins?
"That everyone who wants one and is interested will have their own digital twin from a young age, following them through life," says Mikael Benson.
