TUM’s neuroTUM Student Club Gears Up for Cybathlon, Mastering Brain-Computer Control
The team’s first major international competition, the Cybathlon 2024, is scheduled for the end of October. Representing neuroTUM in the contest is Leon Lucas Joonatan Jokinen, who is unable to move his arms and legs after a swimming accident seven years ago. He will face various virtual tasks using only the power of his thoughts. For example, he has to steer a wheelchair avatar through a room, optimally position a virtual robot arm for an ice cream machine and hold it for a certain time, and find and click an icon on a screen. ‘You need a mental strategy for these tasks,’ says Jokinen. ‘That means focusing on a movement that you can’t perform yourself,’ says the 25-year-old, who enjoys the experience of reactivating the unused areas of his brain.
As a medical student, he is familiar with the capabilities of the human body and the neuroplasticity of the human brain. This is the name given to the ability of brain cells to constantly adapt and, to some extent, to reinvent themselves as needed. The student will be training for several hours a week ahead of the competition: ‘It’s cool to be part of a forward-looking project like this.’
He wears a headset with 24 sensors that measure brain activity using an electroencephalogram (EEG). A computer receives this information, recognizes patterns and derives actions from them. If a person moves their right hand, for example, a certain pattern appears in the EEG, which the computer translates into a movement on the screen. Even if the hand is paralyzed, these regions of the brain can still be activated. A brain-computer interface has been created.
EEG devices donated and supported by Friends of TUM
The neuroTUM team came up with their idea in 2022. ‘We looked at topics we might want to work on and which competitions we could take part in,’ explains electrical engineering student Isabel Tscherniak. ‘At the Cybathlon, the competition with the brain-computer interface sounded very exciting.’ What started as a vague idea has since become very concrete. A manufacturing company donated a portable EEG device and the ‘Friends of TUM’, an association of emeriti, professors and alumni, are supporting the students financially.
As part of neuroTUM, around twenty students from various departments are currently working on clean signal processing, pattern classification with the help of artificial intelligence and machine learning, experimenting and not only making the system real-time capable, but also constantly testing and improving it. ‘Making the systems robust and executable is very complex,’ says computer science student Iustin Curcean. ‘We often spend hours trying to track down and fix errors.’
For the Cybathlon 2024: weekly training sessions with the pilot
Above all, however, preparing for the competition means that the students are investing their time in a project that they have to manage alongside their upcoming exams. ‘We have meetings every fortnight where we discuss the next steps and to-dos,’ explains Isabel Tscherniak. And with the competition just around the corner, there are also weekly training sessions with pilot Leon Lucas Joonatan Jokinen. It is essential for the team to function. ‘It is also important for us to work well together, because, after all, this is a leisure activity,’ says Iustin Curcean. ‘Only if we pull together and are willing to keep at it until after midnight, if necessary, will we be successful in the end.’