#top "top"
Bari, Italy

VIRTUAL BR41N.IO
HACKATHON

‌‌‌‌‌‌ April 17-18, 2021 

during the

Spring School 2021*

 

 

    

 

*BR41N.IO and Spring School 2021 are part of g.tec's Teaching Plan 2021 with more than 140 hours of online courses and lectures.

#winner "winner"

1st PLACE WINNER

Image reconstrucition using GAN technique from EEG signal!

Team members: Lavínia Mitiko Takarabe, Deepak Mewada, Hitendra Singh, Seyedeh Kimia Mousavi, Ardaq Merey, Manvi Jain

 

#video-1 "video-1"

1st PLACE WINNER

NeuroBeat

Team members: Alicja Wicher, Joanna Maria Zalewska, Weronika Sójka,  Ivo John Krystian Derezinski, Krzystof Tołpa, Lukasz Furman, Slawomir Duda

 

#winner "winner"

3rd PLACE WINNER

Multiplay

Team members: Lavínia Mitiko Takarabe, Deepak Mewada, Hitendra Singh, Seyedeh Kimia Mousavi, Ardaq Merey, Manvi Jain

#video-1 "video-1"

1st PLACE WINNER

Top-ECOG

Data Analysis: ECoG Hand-Pose

Team members: Gabriele Penna, Fabrizio Pittatore, Arianna Di Bernardo, Simone Poetto, Simone Azeglio

#winner "winner"

2nd PLACE WINNER

rECoGnise

Data Analysis: ECoG Hand-Pose

Team members: Lorena SantamariaJoao Pereira, Emil Dmitruk, Bartosz Kochanski, Mammad Jamali, Omid Mahdizadeh, Vahid Akbari, Mikolaj Kegler

#video-1 "video-1"

3rd PLACE WINNER

Towards P300 calibration-less single-trial classification

Data Analysis: P300 Speller

Team members: Eduardo Santamaría-Vázquez, Víctor Martínez-Cagigal, Sergio Pérez-Velasco, Diego Marcos-Martínez, Anjali Agarwal, Katherine Rojas

Default Image
#jury "jury"

HACKATHON JURY

 

Image

Christoph Guger (AT)
g.tec neurotechnology GmbH

 

Image

Dean Krusienski (US)
Virginia Commonwealth University

 

Image

Grace Rigdon (US)
IEEE Brain Initiative

 

Image

Abdelkader N. Belkacem (AE)
United Arab Emirates University

 

Image

Natalie Mrachacz-Kersting (DE)
Dortmund University of Applied Sciences and Arts

Image

Milena Korostenskaja (US)
 Florida Hospital for Children

 

BCI PRINCIPLES

 

Brain-computer interfaces are realized by 4 different principles:

slow waves

steady-state visual evoked potentials (SSVEP)

motor imagery (MI)

evoked potentials (EP)

In the BR41N.IO Hackathon Series, motor imagery and EP based systems
are mostly used to control the applications:

 

In the case of the motor imagery application, participants have to imagine e.g. left or right hand movement to produce an event-related desynchronization over the sensorimotor cortex. This is basically an amplitude change of the alpha and beta regions of the EEG.

In the case of EPs, the BCI system is showing different flashing icons and the user has to attend to the icon he wants to select. When the icon flashes on the computer screen, than a P300 wave is produced in the brain and the BCI system is able to detect it.
 

#locations "locations"

ON-SITE LOCATIONS

 

G.TEC AUSTRIA

g.tec neurotechnology GmbH
Schiedlberg, Austria

UAEU ABU DHABI

  United Arab Emirates University
  Abu Dhabi, United Arab Emirates

MLJC ITALY

Machine Learning Journal Club
Turin, Italy

#projects "projects"

PROGRAMMING PROJECTS

 

Unicorn Speller: Smart Home

The Unicorn Brain Interface comes with the Unicorn Speller application that is using P300 paradigm to control electronic devices such as lamps, radios or television. Watch the video Unicorn Speller Smart Home.

soft-/hardware: Unicorn Hybrid Black, Unicorn Speller, electronic devices
participants: 3-5 people per group
skills: basic programming skills (Matlab, Simulink)

 

‌‌‌ Orthosis Control

It is possible to control a 3D printed orthosis using a Unicorn Hybrid Black with motor imagery. It is possible to move an orthosis by thinking about left or right hand motion. Watch the Orthosis Control video.

soft-/hardware: Unicorn Hybrid Black, Unicorn Speller, bring your own orthosis or use Unity instead
participants: 3-5 people per group
Skills: Basic programming skills (Matlab, Simulink), basic graphics programming skills with Unity

 

Unity Rehab

Create a Unity based game that can be used for rehabilitation purposes.

soft-/hardware: Unicorn Hybrid Black, Unity
participants: 3-5 people per group
Skills: Basic programming skills (Matlab, Simulink), Basic graphics programming with Unity

 

‌‌ Dream Painting

If you want to create a Dream Painting,  you have to wear a Unicorn Brain Interface while you sleep. When you wake up, you are able to create a picture based on EEG signals.

soft-/hardware: Unicorn Hybrid Black, Unicorn Painting
participants: 3-5 people per group
skills: Basic programming skills (Matlab, Simulink)

 

‌‌‌ fNIRS and EEG Control

The team can use fNIRS (functional near-infrared spectroscopy) and EEG simultaneously to control BCI applications.

soft-/hardware: g.Nautilus fNIRS
participants: 3-5 people per group
skills: Basic programming skills (Matlab, Simulink)

 

 

‌‌ Unity Games

Create your own Unity game that can be controlled with a brain-computer interface.

soft-/hardware: Unicorn Hybrid Black, Unicorn Suite, Unity
participants: 3-5 people per group
Skills: Basic programming skills (Matlab, Simulink), Basic graphics programming with Unity

 

 

‌‌ Unicorn Sphero

The Unicorn Hybrid Black offers the Unicorn Speller application that allows you to control a robotic ball called Sphero. Watch the video Unicorn Sphero.

soft-/hardware: Unicorn Hybrid Black, Unicorn Sphero, Sphero robot
participants: 3-5 people per group
skills: Basic programming skills (C#)

 

‌‌ Flight Control

The Unicorn Hybrid Black comes with the Unicorn Speller which allows you to fly your own drone with the brain only.

soft-/hardware: Unicorn Hybrid Black, Unicorn Speller, bring your own drone or use Unity instead
participants: 3-5 people per group
skills: Basic programming skills (Java)

 

Your Hacking Project

You are invited to create your own programming project for this hackathon. You'll have all the BCI headsets or you bring your own BCI to design and program your own fully functional headset.

soft-/hardware specifications: Unicorn Hybrid Black, tbd
participants: 3-5 people per group
skills: Basic programming skills

 

 

OFFLINE DATA ANALYSIS PROJECTS** 

**no hardware required.
 

‌‌‌ Stroke Rehab Data Analysis

Analyze a motor imagery BCI data-set from a chronic stroke patient in order to optimize pre-processing, feature extraction and classification algorithms. Compare your results with state-of-the-art algorithms.

soft-/hardware: MATLAB or other signal processing platform that is able to read in the MATLAB matrix
participants: 3-5 people per group
skills: signal processing skills

 

 

‌‌ P300 Speller Data Analysis

Analyze a visual P300 BCI data-set from a healthy person in order to optimize pre-processing, feature extraction and classification algorithms. Compare your results with state-of-the-art algorithms.

soft-/hardware: MATLAB or other signal processing platform that is able to read in the MATLAB matrix
participants: 3-5 people per group
skills: signal processing skills

 

 

Unresponsive Wakefullness Syndrom Data Analysis

Analyze a vibro-tactile P300 BCI data-set from a patient with disorders of consciousness in order to optimize pre-processing, feature extraction and classification algorithms. Compare your results with state-of-the-art algorithms.

soft-/hardware: MATLAB or other signal processing platform that is able to read in the MATLAB matrix
participants: 3-5 people per group
skills: signal processing skills

 

‌‌ SSVEP Data Analysis

Analyze an SSVEP BCI data-set from a healthy person in order to optimize pre-processing, feature extraction and classification algorithms. Compare your results with state-of-the-art algorithms.

soft-/hardware: MATLAB or other signal processing platform that is able to read in the MATLAB matrix
participants: 3-5 people per groupskills: signal processing skills

 

Locked-in Patient Data Analysis

Analyze a vibro-tactile P300 BCI data-set from a patient with locked-in syndrom in order to optimize pre-processing, feature extraction and classification algorithms. Compare your results with state-of-the-art algorithms.

soft-/hardware: MATLAB or other signal processing platform that is able to read in the MATLAB matrix
participants: 3-5 people per group
skills: signal processing skills

 

 

‌‌ ECoG Hand Pose Data Analysis

Analyze an ECoG BCI data-set from an epilepsy person in order to optimize pre-processing, feature extraction and classification algorithms. Compare your results with state-of-the-art algorithms.

soft-/hardware: MATLAB or other signal processing platform that is able to read in the MATLAB matrix
participants: 3-5 people per group
skills: signal processing skills