Students work on their thesis for weeks, months, sometimes years. Once completed, their work is often relegated to a dusty drawer. The Thesis Tales column is devoted to highlighting student research. In episode 3: Femke Aminetzah (21, Bachelor’s in Artificial Intelligence) investigated the best way to reduce motor symptoms in parkinson’s patients using AI.
Can you tell us in a few sentences what your thesis is about?
‘I investigated how AI can help alleviate motor symptoms in Parkinson’s patients. Parkinson’s disease is a condition in which dopamine producing brain cells are broken down, leading to various motor and psychological symptoms. I focused specifically on motor problems and the possibilities of improving these using AI. An existing treatment method is Deep Brain Stimulation (DBS), but it is often associated with side effects. In my research, I looked at whether AI could help reduce these side effects.’
‘I examined four algorithms designed to analyse brain activity’
What exactly is Deep Brain Stimulation?
‘In this treatment, an electrode is placed in the brain at locations where dopamine-producing cells have decreased. This electrode delivers electrical stimulation, mimicking the effect of dopamine. The problem with DBS is that it often causes negative side effects, such as depression. This is because constant dopamine-like stimulation is not always necessary, and can thus lead to over-stimulation of certain brain regions.’
What exactly did you do in your research?
‘I have developed a DBS device that can read the activity in the brain and, based on this, determine how much mimicked dopamine is needed at any specific moment. I did this by examining four existing algorithms designed to analyse brain activity. Using a computer simulation, I tested which algorithm is best suited to accurately estimate the required dopamine. Based on the algorithm that came out as the best, I drew up a stimulation plan for the DBS device.’
Why did you choose this topic?
‘I wanted to do something clinical because I think AI should be applied more in the medical sector. There is a lot of fear surrounding the use of AI in healthcare. Although the risks of AI-driven medical devices can be high, they often perform well, are efficient, and can reduce costs. Yet, promising research on these technologies often remains unused due to strict legislation. This has to change.’
‘I had to take responsibility for my own ideas and provide a lot of supporting evidence for them’
How did you experience the process of writing your thesis?
‘It took me a long time to get certain codes to work. This meant that I ran out of time and even had to work through the summer holidays. It bothered me that I could not move forward as planned, but it improved my programming skills. In addition, I had a different approach than my supervisor. As a result, I had to take responsibility for my own ideas and provide a lot of supporting evidence for them. This ultimately gave me a lot of confidence. Once my code finally worked and I could start working with the results, the process suddenly went very fast. In the end, I was very proud of my thesis.’
What can students learn from you?
‘It is really important to choose something you are interested in. That helps you pull through. When supervision is not going well or something is not working out, a fun topic can give you the motivation to keep going.’
Great that you are reading Vox! Do you want to stay up to date on all university news?
Thanks for adding the vox-app!
