| News | Interview | “A kind of organ donor card for data use can facilitate innovation in medical technology”

“A kind of organ donor card for data use can facilitate innovation in medical technology”

From local start-up to world market leader in software-based medical technology - Stefan Vilsmeier founded Brainlab over 30 years ago. The employees develop innovative software and hardware solutions in the fields of surgery and radiotherapy. An interview about data use for life-saving medical technology and the question of whether there should be something like an organ donor card for the use of data.

CEO Stefan Vilsmeier in conversation with bidt Science Communication Manager Nadine Hildebrandt at the Munich headquarters of Brainlab AG.
© bidt / Kilian Blees

What is important in the digital transformation?

Stefan Vilsmeier: In medical technology, three pillars are crucial: firstly, the technology itself – i.e. the software and hardware; secondly, access to data; and thirdly, the change in people’s minds. It was clear to me very early on that the third pillar in particular – and associated with it, the provision of information and training on new technologies – is fundamental. For this reason, back then, as a still very small company, we invested in education and training, for example, and today training or even hackathons are an essential pillar.

What does it mean to digitise surgery?

You want to record the position of an instrument precisely – but also understand how the whole thing interacts with the patient. Our medical technology works like a GPS: Where is the device located during an operation? Which structure is on the left, and which is on the right? What happens if the attending doctor goes down three millimetres? We have mapped the patients and combined diagnostic data from computer and magnetic resonance tomographies, for example, in a model. For example, implants can be placed precisely in an accident patient.

AI, VR and robotics – can new technologies replace the doctor?

Today, the use of artificial intelligence, virtual reality and robotics opens up completely new possibilities than was the case ten years ago. However, some people think that if the robot does it, everything is much better, especially in the field of robotics. But at the beginning, you must consider: What problem do I want to solve in concrete terms? Will the operation be faster, safer, cheaper and gentler for the patient through the use of robots? Sometimes precisely the opposite is the case. All the robots are rather an assistance system that supports the surgeon, for example, by having the robot arm hold something during the operation. The same goes for AI – it can help analyse CT scans or X-rays and give a preliminary diagnosis. However, medical professionals still bear the responsibility and decide.

Stefan Vilsmeier veranschaulicht die Einsatzmöglichkeiten von Virtual Reality in der Medizintechnik im Brainlab-Headquarter.
Stefan Vilsmeier illustrates the possible uses of virtual reality in medical technology at Brainlab headquarters.

You mention the use of data in medical technology – what needs to be taken into account here?

Transparency is enormously important here, and with it, the question is: who does what with my data and where? I am speaking here as an entrepreneur but also as a citizen and potential patient – and of course, I want to feel comfortable with what happens with my data. I believe data should also be distributed as widely as possible, i.e., not one central state authority watches over all the data. Still, I see great opportunities in a distributed decentralised system with many actors.

Standardised and transparent interfaces are important. Good approaches are those that put citizens at the centre.

Stefan Vilsmeier, CEO Brainlab AG

What can the use of data in medical technology achieve?

In Parkinson’s disease, for example, mechanical targeting devices can be used to place an electrode in a specific area of the brain. Suppose we collect data from a sufficient number of patients. In that case, we can much better identify the so-called sweet spot for deep brain stimulation – in other words, the area where targeted stimulation is needed. If this point can be found better, thanks to a lot of patient data, then the duration of the operation is reduced from eight hours previously to one hour – the stress for the patient is noticeably reduced.

To what extent does the diversity of the data play a role?

You always need data of excellent and comprehensible quality. But obtaining data is not always easy – especially in clinical trials, there are often very narrow criteria – and these may not be chosen broadly enough. It is also essential that data from many different people, for example, in terms of age or gender, are incorporated into the development of new technologies. If, for example, an algorithm for detecting skin cancer is only trained with data from light-skinned people, then it does not reliably detect possible cancer stages on darker skin, and misdiagnoses occur. This is an example where you have to work specifically on data quality.

You always need data of excellent and comprehensible quality.

What role do data donations play?

Up to now, the data we use are on the systems of our users – i.e. the clinics. They are managed by them and used for primary purposes. We, as manufacturers, do not have access to the data at first. The great opportunity, of course, would be the secondary use so that not only the specific illness of the individual patient can be optimally treated but also that actors outside the hospitals can advance innovations – for example, for the treatment of cancer. For example, the BORN project of the Bavarian Centre for Cancer Research is currently exciting: here, for the first time, the collection of health data is standardised, structured and process-oriented right from the start to create a globally unique data basis for the development of image-based biomarkers and AI machine learning methods.

How could data use be simplified here?

First of all: It is good and right that data can only be collected with the appropriate consent of the patient. However, during a hospital stay, patients are currently expected to do a lot of paperwork about the corresponding consent. Many pages must be read through and signed on the spot – the patient can quickly lose the overview – this also applies to revocation options.

In my view, many things could be simplified here; for example, by introducing a data release passport for consent to data use, as with the organ donor card.

Stefan Vilsmeier, CEO Brainlab AG

What should be taken into account?

Basic questions should always be patient-centred: What do I want to allow as a patient? Who can process which data for which purpose and with which access method? There are already promising approaches, such as the medical informatics initiative in the form of a uniform declaration of consent for the broad use of patient data. Unfortunately, the problem with this is that the consent is not very comprehensible. And from an entrepreneurial point of view, I have to say that this does not explicitly address the industry as a user. We have therefore become active ourselves – for two years, we have been in contact with patient organisations, politicians and data ethicists, lawyers, and philosophers and are developing our template for consent. The exchange with many actors from research and science and transparency are important; we only want to include data in our research that is explicitly made available to us. But it is also becoming apparent that, in addition to software and hardware development, the design of data use is a central success factor for digital transformation.

Thank you very much for the interview!

The interview was conducted by Nadine Hildebrandt.