The goal for this project would be to create a new generation of portable Brain Computer Interface (BCI) that are compatible with the Raspberry Pi.
The commercially available BCIs can collect data up to 250 Hz and there is not a guide to use the BCI on a Raspberry Pi 4. Also, they cost more than 1000 euros. For my Senior Design Project, I designed a Smart Wheelchair that runs with a BCI, Unicorn Hybrid Black. People who suffer from locked-in syndrome can't move their limbs to control the joystick of a motorized wheelchair, so the idea of the Smart Wheelchair was that a BCI reads the intention of the user and activates servo motors to move the joystick towards the desired location. By using a Raspberry Pi 4, we could implement an embedded system and easily fit into the design of the Smart Wheelchair. The major problem I had with the project was all the time I spent installing open source software and the low frequency sample.
There are new open-source boards (PiEEG and HackEEG) that claim they can sample data up to 16 kHz, can be easily installed either on a Raspberry Pi 4 or on an Arduino, and can cost as low as 250 euros. The high sampling rate allows for better frequency analysis of brain waves and allows to collect higher brain wave frequencies that were not possible to collect with the previous generation of commercially available BCI. These boards are shields so they can't work independently and this seriously limits the design for portable BCIs.
PiBCI will be the first full BCI to collect EEG up to 16 kHz and be easily installed on a Raspberry Pi. Open source designs from OpenBCI and PiEEG can be used to design a portable BCI that works autonomously and sends data to a Raspberry Pi 4. This project will re-design the concept of BCIs in order to be almost invisible during operation and to easily apply all the electrodes on the scalp. Analyze higher frequencies during every-day activity and their associated role. In summary PiBCI's project will:
- Design novel BCI that can be easily installed on a Raspberry Pi 4
- Collect data up to 16 kHz
- Design cheaper and more neuroergonomic BCIs
- Analyze higher brain wave frequencies
- Select the most neuroergonomic and with the highest signal to noise ratio electrodes.
The first month of this project will be dedicated to designing the PCB and new neuroergonomic designs. Eagle can be used to design the PCB since it is a really popular tool and it enables easy 3D designs. *** can manufacture the PCB in just a few days. In the meantime, I can 3D print the shield of the PCB and study open-source code and API to collect data from the PCB. The rest of the second month, I can spend it by connecting the BCI to the Raspberry Pi 4 and stream data. During the final month, I can do some Motor Imagery and P300 tasks to analyze brain waves and compare it with other available data.