Improving modern-day medicine using Data Science
Technology has changed both the medical and data science field a lot over the last decades. As data science is still a new field, it has lots of potential to bring big changes within the medical domain. To illustrate that data science research can improve medical practice, D.S.A. Pattern is organizing a Datathon together with the ‘Elisabeth-Tweesteden Ziekenhuis’ (ETZ) in Tilburg. Participants will be using anonymized data from the Department of Neurosurgery at the ETZ. We interviewed Geert-Jan Rutten, a neurosurgeon at the ETZ, to find out how efforts regarding data science can improve the field of medicine.
How got Geert-Jan, a neurosurgeon, interested in data science research? He explains that his interest in science and technology already started when he studied physics at Utrecht University, before starting his education in Medicine. Later, he did his Ph.D. in functional MRI in neurosurgical patients to predict the location of motor and language in the brain. During his education he had an internship at the ETZ which lead to his permanent function. At the ETZ, Geert-Jan is focused towards awake brain surgery and brain-stimulation for patients with Parkinson’s disease. As a neurosurgeon, his research interest focuses on the cognitive abilities of brain tumor patients.
In the interview, Geert-Jan mentions that traditionally, neurosurgeons used to focus on neurological functions such as movement and speech, as a proxy for quality of life for the individual patient. What is much more important, in fact, is to know a patient’s individual goals in life, as well as his or her socio-professional status. Only then can we try to maintain a person’s quality of life after complex brain surgery. However, this information was largely missing. 10 years ago, the Department of Neurosurgery of ETZ together with the Department of Cognitive Psychology of Tilburg University started to systematically test the cognitive functions of patients before and after surgery. It was discovered that there were much more problems than expected with cognitive function (years) after surgery.
“Not being able to concentrate well, or fatigue is what really diminishes your functionality and your ability to return to a normal social and professional life.”
When a patient has a brain tumor, there are usually multiple treatment plans. The doctor has to decide, together with the patients, what the best possible treatment plan is, balancing optimal cancer treatment with optimal preservation of neurological and cognitive functions. To date, it is not yet possible to predict cognitive outcomes after surgery of each individual patient. Current knowledge is based on results in the general (brain tumor) population. However, there are marked individual differences, as every person reacts differently to treatments and surgery. Therefore, studies are currently underway to investigate personal outcomes for patients.
How can data science and in particular a Datathon bring a solution to this problem? The goal of the Datathon is to predict cognitive function from functional MRI data. How does this help improve the way medicine is practiced? Geert-Jan knows that with a Datathon the chances of finding an ideal solution are not very high. However, a lot of groups working on a challenge can bring new valuable insights in how to handle the provided data.
The ideal end goal would be to know the long term effects of surgery on a patient by using multiple data sources and that way being able to fully simulate a surgery and its outcome. That is an optimal end goal which is still a spot on the horizon, but linking data from functional MRI scans to cognitive function in individual patients is already a small piece in the big puzzle.
The Pattern Datathon will take place on the fourth of June 2021, at Tilburg University. More information and sign-ups can be found at dsapattern.nl/datathon2021. Sign-ups are open until Wednesday the 26th of May and can be done either individually or in teams of four. For questions about the event contact us at datathon@dsapattern.nl.
