Site Under Construction
Welcome to BCI cellar. My name is Wal Haywood and I’m experimenting with Brain-Computer Interfaces (BCI). As you may have guessed, the project work is carried out in my cellar. I live near the English Lake District and people working in this field are a bit thin on the ground in this fairly remote location. The purpose of my website is to share what I have done so far and, hopefully, to communicate with others working on similar projects.
If you are new to BCIs or EEGs and are wondering what this is all about, have a look at the introductory page About BCIs
The EEG equipment I am using was purchased from OpenBCI. It comprises a Ganglion 4-channel board and an EEG electrode Cap Kit with sintered electrodes which use an electrode gel. This is my most recent headwear. I have tried the Ultracortex headset and also gold cup electrodes, but I’ll put the nine-year history of the project on a separate page.
The aim of the project in its current form is to try to get reasonably distinguishable variation in processed signals from the electrodes in response to motor imagery. That means imagining movement on the right or left side of the body, or indeed stillness, and monitoring and recording the EEG outcomes to look for patterns.
My software is written in Unity and C# and it uses the BrainFlow software for signal acquisition and processing. A video of the software in action is shown below but it will be explained in detail on other pages. Please ignore the actual numbers in the video, I was too busy capturing the screen to worry about imagining movement!
I have used a biofeedback approach to see if the brain could be trained to produce a visible response on the screen. This is in much the same way that biofeedback can be presented to aid relaxation by showing the strength of alpha waves. I am using expanding and contracting circles to represent the strengths of beta and mu waves at the electrodes selected. You can see from the screen that I am monitoring the C3 and C4 positions above the motor cortex, and also P3 and P4 above the parietal cortex. I may change these electrode positions in a future version since P3 and P4 may not be the most useful for motor Imagery, however I am constrained by available positions in my EEG cap. Details of the software are on the Project page. I am using these biofeedback principles to see whether the system can detect any bilateral asymmetry in brain activity at the selected frequencies in response to purposefully imagined movement of one side of the body or another
I create logs which record results and therefore give me some idea of progress. If my approach is even remotely successful, in the longer term I will try to train an AI to recognise the side which is imagined moving from the responses produced.
If this introductory summary has piqued your interest, then read on.