About

Working towards restoring autonomy.

We’re neuroTUM, a non-profit organization working on neurotechnology to improve the independence of people with physical disabilities.
‍
According to the WHO between 250 000 and 500 000 people become paralyzed each year losing all forms of mobility which most of us take for granted. In severe cases communication with the outside world is no longer possible.

Brain-computer interfaces (BCI) can help restore the autonomy of affected patients by decoding their neural activity into precise control signals which can control a wheelchair or be translated into speech.

To ensure the reliability of our BCI, our first goal is to control a car in a virtual environment by non invasively recording electrical brain signals using an Electroencephalogram (EEG). We will present the developed technology at the Cybathlon BCI competition in 2024. Here, internationally renowned research groups compete together with talented paraplegic pilots in a virtual race. The research groups provide their pilot with a BCIs to be used by their pilots for steering a vehicle with the pilots neural activity as only control signal. The goal of this project is to establish the technology which can be subsequently transferred to the reliable control of a wheelchair.

Join us on our mission to develop a more inclusive future!

Organization of the Club

Our Subteams

Experimental Design

In this group we plan and execute our experiments. For our current BCI (brain-computer interface) project this includes designing the instructions we give to our subjects in order to evoke distinguishable neurological responses. This enables the signal processing and deep learning team to reliably classify the pilot’s intention. It is also in our responsibility to mimic the test scenario of the BCI-Cybathlon competition by understanding their regulations and developing a test environment.

Signal Processing

The neural activity of our pilot is measured using an EEG (electroencephalogram) and fNIRS (functional near-infrared spectroscopy). The data collected from our EEG and fNIRS is filled with noise, artifacts of involuntary movement and activity of brain areas which are only weakly involved in the desired task. The signal processing team is creating a pipeline to transform the raw EEG and fNIR data into high information dense and relevant subsets that can subsequently be analyzed by the deep learning team.

Deep Learning

The task of our team is to classify the pre-processed measurement data into distinct control signals to use as feedback in real time. To achieve this, we create our own pipeline implementation and classify our data with models such as LSTM (Long Short-Term Memory), more recent approaches like transformers and other state-of-the-art machine learning methods that we will implement with custom solutions for BCI.

Public Relations

The PR team is responsible for creating an identifiable brand for our club through the logo design and color choices. Additionally, they are responsible for designing and maintaining the website and social media pages. Furthermore, this team will be establishing and maintaining contact with partners and sponsors. In future, they will also organize public educational events.