My first encounter with academia
I started my academic career at the Radboud University (RU) in 2019. After considering the Computer Science and Artifical Intelligence, I decided to opt for the latter, since it had a numerus fixus.
This turned out to be a great choice for me, as I fell in love with the field. I especially enjoyed how the RU emphasised that AI lies at the intersection of computer science, neuroscience, psychology, and social science (think HCI). I quickly came to find that courses such as Brain, Cognitive Computational Neuroscience, and Cognitive Neuropsychology really caught my interest, next to the more 'traditional AI' courses like Deep Learning. When given the choice of free electives, I then found myself choosing a lot of psychology and neuroscience courses.
The thesis: Neuroscience 🤝 AI
This also steered my thesis subject to the neuroscience-meets-AI corner. Under supervision of Jordy Thielen, I worked on decoding EEG brain signals. More specifically, deciphering what participants saw by reading out EEG signals from the visual cortex.
What I also enjoyed was that the application of this research was easily defendable: to allow people who can't move anymore to continue to communicate. The participants could look at the keyboard on the screen, and by decoding the (very noisy) EEG data, we could discover the respective flashing codes. Since each letter flashed with a different code (pattern of periods of black and white background), the procedure is called code-modulated visually evoked potentials, c-VEP.
In my thesis I explored ways to improve the accuracy and speed of typing in the experimemt. Sadly, my attempts did not provide significant improvements. The research did however finally force me to understand statistics.
In the end, I achieved an 8/10 for the BS thesis on improving the performance of Brain Computer Interfaces using Code-Modulated Visually Evoked Potentials (c-VEP).
Find the thesis here.