Signals & Projects

What real IronBCI-8 recordings look like, open data to practise with, and projects to build on. All recordings below were made with dry Ag/AgCl electrodes, no gel, at 250 samples per second.

Example signals

EEG: alpha rhythm

Eyes open and closed in turn. The orange line is high while the eyes are closed. The alpha rhythm grows in those periods, and the wavelet view (bottom) shows the extra power in the alpha band.

Alpha rhythm test with IronBCI-8 and wavelet transform

EEG: chewing and blinking artifacts

Raw data. The large bursts are chewing and blinking: muscles and eye movement, not the brain.

Chewing and blinking artifacts recorded with IronBCI-8

Learn to recognise artifacts before you analyse anything. Remove or mark them before computing band power or training a model.

Dataset

In the dataset (opens in a new tab) folder of the IronBCI repository. Alpha test: 5 seconds eyes open, 5 seconds eyes closed, repeated. 250 SPS, dry electrodes without gel.

File	Content	Size
`eeg_raw_Data_alpha_test.xlsx`	Raw data in µV	4 channels, about 50,000 samples (200 s)
`eeg_band_pass_filtered_alpha_test.xlsx`	The same data after an 8–12 Hz band-pass filter	4 channels, about 50,000 samples

To plot your own recordings with a band-pass filter, use 1.Band_pass_filter_data_visualisation.py (opens in a new tab). Set the file name and the filter range at the top of the script.

Projects

Real-time emotion detection with ChatGPT

Chat_GPT_Real_time.py (opens in a new tab) streams EEG from IronBCI-8, computes alpha-band power every second, and asks an OpenAI model to label it as Scary, Sad or Funny, using a few labelled examples in the prompt.

pip install bleak numpy scipy pandas openai
python Chat_GPT_Real_time.py

Set your own OpenAI API key first. Store it in an environment variable rather than in the script. Step-by-step guide: Connect Brain to ChatGPT (opens in a new tab).

⚠️

This is a fun learning example. A few labelled alpha-power values are not a validated measure of emotion.

3D-printable headset

An open, 3D-printable EEG headset designed for IronBCI, for DIY neuroscience, teaching and BCI experiments. You can adapt it to your own electrode layout.

Files and instructions: ironbci_3D_EEG_Printable_Headset (opens in a new tab).

Build on the app

The Android app is open source (Flutter, MIT licence). Start from EEG-Flutter-App-master.zip (opens in a new tab) to build your own mobile BCI app.

Software Specification