Hackathon - BR41N.io 2024

Summary

In this hackathon hosted by g.tec medical engineering, we had 24 hours to build something cool with gtec’s brain-computer interface technology. My team and I worked on a dataset provided to us by the hosts, which included EEG recordings of stroke rehabilitation patients, who were asked to imagine moving their left or right hands. We implemented and evaluated some classifiers to detect either left hand or right hand imagined movement, from the recorded EEG.

Skills Gained

• Collaboration with people you just met, from all around the world!
• EEG signal processing methods for real-time signal detection
  • Noise removal (bandpass filtering)
  • Feature extraction (common spatial patterns)
  • Classification (linear discriminant analysis)
• Evaluation of signal processing pipeline performance, assessing accuracy, computation time and generalisability

Code

Repository: neurosparks-gtec-hack

This repo also presents our findings, after explaining our methods.