NeuroPi

Thank you for visiting the NeuroPi repository. This Python code is designed for use with the OpenBCI interface on Raspberry Pi 3.

Table of Contents

• Python Libraries and OpenBCI
• Setting Up Raspberry Pi
• Code
• Datasets

Python Libraries and OpenBCI

This repository contains the necessary code and instructions for setting up a Brain-Computer Interface (BCI) system using Python and OpenBCI. The required Python libraries can be installed via PyPi. To install these libraries, you can use the following terminal commands:

pip3 install matplotlib
pip3 install numpy
pip3 install pylsl
pip3 install python-osc
pip3 install pyserial
pip3 install requests
pip3 install socketIO-client
pip3 install websocket-client
pip3 install wheel
pip3 install Yapsy
pip3 install xmltodict
pip3 install scipy
pip3 install openbci-python

Setting Up Raspberry Pi

The ARM architecture does not compile the file liblsl32.so which is present in the pylsl library, responsible for data transmission. Follow these steps to replace the file:

1. Download the replacement file (liblsl-bcm2708.so).

2. Unpack the downloaded file.

3. Use the following terminal command to replace the file:

   mv <path_to>/liblsl-bcm2708.so <pylsl_path>/liblsl32.so

Please replace link_to_download with the actual URL to download the replacement file.

BCI System README

Code

Download the necessary code from this link. This folder contains all the required files for the BCI system. The "offline" folder does not include the code for the 16-electrode online system. To set up and run the complete system, use the "stream_online.py" code located in the "online/scripts" directory. This code orchestrates other functions. If you need to change essential information (acquisition count, electrode count, etc.), make changes in the "main.py" and "stream_time.py" programs.

• stream_online.py: Responsible for online data acquisition and classification, executing "main.py" and "stream_time.py".
• main.py: Manages offline data windowing, filtering, and classification.
• stream_time.py: Handles data acquisition and prepares the training dataset.

Control variables for the BCI in the "stream_online.py" program have values of 1 or 0:

• treino: Enables system acquisition to generate training data (used in "stream_time.py").
• aplica_filtro: Applies filtering and performs offline BCI with or without attribute selection (used in "main.py").
• classifica_online: Conducts online BCI acquisitions based on the specified number.

Adjust the runfiles paths in the "stream_time.py" code if necessary, indicating the locations of "main.py" and "stream_time.py". Configuration changes (frequency count, electrode count, acquisition count, etc.) should primarily be made in these three programs using variables at the beginning of each.

Datasets

The system offers 8 datasets (four with 8 electrodes and four with 16 electrodes), both with original and filtered signals. Directories for each acquisition are as follows:

• /online/scripts/Data_record/V_1_16
• /online/scripts/Data_record/V_2_16
• /online/scripts/Data_record/V_3_16
• /online/scripts/Data_record/V_4_16
• /online/scripts/Data_record/V_1_8
• /online/scripts/Data_record/V_2_8
• /online/scripts/Data_record/V_3_8
• /online/scripts/Data_record/C_1_8