THE ONLY GUIDE YOU’LL EVER NEED TO UNDERSTAND BCIs
Become Godly With Your Brain
✨Recommended Article: https://sayyidaraniahashim.medium.com/meet-your-brain-1e749cd0860
Have you ever come across the popular myth, “you only use 10% of your brain”?
Being the type of person obsessed with “potential”, I remember sitting up straight, giving my undivided attention to what followed this quote. I was under the impression that I could harness this to increase my intelligence — which in hindsight, is a very unintelligent thing to believe.
A lot of times, people end up believing that the rest of the 90% could be used to get godly powers like telekinesis.
🚨NEWS FLASH: YOU USE WAY MORE THAN 10% OF YOUR BRAIN… BUT THAT WON’T STOP YOU FROM BECOMING A GOD 🚨
“Wait… Repeat the last bit again?”
You read that correct, telekinesis IS possible — but it isn’t some spell or spiritual energy. It is powered by your brain (and a very special technology) 😉
Excited? Well, save some for later because you’re going to be reading about some really cool things — which could soon make you:
Okay fine, you’ll have to wait a tiny bit more for that — but I promise you that you’ll be finishing off this article spellbound.
How do I stop a football from causing me major injuries?
The answer: Brain Computer Interfaces.
Know your basics.
Brain-Computer Interfaces is a tool or a system that allows your brain to directly communicate with a machine, using brain signals! You’re basically controlling the world with your 🧠
This is usually done via 3 steps, namely collection of brain signals, interpretation and analysis and finally, translation of these signals into the corresponding commands (AKA stopping the football) from the interpreted data.
There are many types of brain signals, which can be used to spit out our output.
The human brain is incredible. Containing about 86 BILLION neurons linked to each other, they release little electrical signals which move from neuron to neuron that make you do things like have an existential crisis.
All the electrical signals fall in either of two categories:
- Field Signals
There’s one problem BCIs face, though: the electrical signals are weak and variable, making them harder to use. Artificial Intelligence/Machine learning is the most efficient tool for BCI systems because of their accuracy in solving problems like this.
How are Electrical Signals Formed?
Now, this all relates to the “membrane potential”, “action potential” stuff you read about earlier. If you haven’t read that article yet — check it out here! It’ll make understanding electrical signals SO MUCH easier:
Meet your Brain!
An overview of neurons, membrane potential & the like. the Part 1 of “Lighting Up Your World” article (about…
The Many Steps To BCIs
Earlier, I described the three basic steps to make the BCI magic happen. Let’s look at them in-depth:
- Collection Of Brain Signals
This step includes the brain signal production as well as the detection aspect.
“Production? Isn’t that happening, like, every second?”
If you had the above thought — good job! You’re clear with the concept.
Anyway, to answer the question, we know that with every thought and action, our neurons are firing synapses. This is known as spontaneous brain activity. Now, we could just use these signals that are already generated by the subject.
However, if we are looking for a specific type of brain signal, we’ll need to trigger it. The way we do this is by presenting stimuli to the subject or having the subject imagine specific things. We call this induced brain activity. Actively generating signals makes detection easier by giving you control over the stimuli.
Whatever method you use, one thing that’s STILL missing is detecting these signals. I mean, what’s the point of having all these signals flowing through you if the system isn’t detecting them?
This is where signal detection comes in.
While there are multiple ways of detection [both non-invasive, invasive AND semi-invasive], the most common ones are EEG and ECoG.
EEG is short for electroencephalography.
Alright, it might be a bit of a mouthful of a word, but I promise you it isn’t difficult at all to understand.
EEG is essentially a device used to measure the brain’s electrical activity, more specifically the activity of large groups of neurons in the cerebral cortex.
It usually measures postsynaptic potentials. Now, this isn’t the same as action potentials, but it is related. When this become excitatory, the neuron is triggered to release an action potential.
Note: For electrodes, the policy is “the more, the merrier.” A minimum of two electrodes are required as the EEG is essentially the difference in voltage between electrodes.
As of for how this is all done, the EEG machine has three main components:
- Electrodes: These small metal discs are the stars of the show! These electrodes, as you know, pick up electrical signals and may be combined to an electrode cap. There are also different types of electrodes; wet, dry, active and passive. Wet electrodes are generally made of silver/silver chloride material and use electrolytic gel material as the conductor. Dry electrodes consist of a single metal acting as a conductor between the skin and the electrode. About active and passive electrodes — you’ll need to learn a bit more to understand that, so follow along to the next section!
- Amplifier: As EEGs are non-invasive, signals are usually poor. The EEG amplifier converts the weak signals into a more detectable signal to help us understand. This is the differentiating point between active and passive electrodes. Active electrodes have a pre-amplification module immediately after the conductive material unlike passive ones which simply extends the connection from the conductive material to the equipment. Active electrodes are better in terms of quality (as there will be a reduced amount of noise), but cost more and are heavier.
- Analog to Digital Converters: For the data to be processed and understood by a computer, the data needs to be converted from analog to digital. This is done by the A/D converters.
This is what you find in EEG reports:
Notice the differences in the waves?
Yeah, the waves can be segregated into different types based on factors like frequency (of waves) and amplitude (height and depth of the waves). Generally, these are of 5 types; Alpha, Beta, Gamma, Delta and Theta.
All of these waves are indicative of many things and constant monitoring can give us some pretty cool insights.
Brainwaves; Our Brain’s Electrical Messages
While we cannot generalise the meaning of the waves as they change in location, they do open to us a lot of hints 🔑
So, without further ado, let’s dive into these!
By the look, the name and the speed, you get the vibes that these are calm, serene and dreamy. I mean, I’m getting summer-beach-vibes just by looking at the animation below!
Delta waves are characterised by a very high amplitude (greatest among the waves) and very slow frequency. Deep, dreamless sleep takes you down to this type of waves.
This usually centre around 1.5–4 cycles per minute, but never go to 0 (because that’d mean that you are braindead!).
Too much delta waves could potentially mean that you have a learning disability, severe ADHD, brain injuries or an inability to think, while too little delta waves could mean poor sleep or an inability to rest and rejuvenate.
Now, a lot of you probably have bad experiences with the word “theta” because of your high school trigonometry classes ➗, but bear with me because these types of waves are really important.
Anyway, these usually have a smaller amplitude than delta waves and faster frequency. They are typically associated with daydreaming.
If you daydream often, you know that it’s *THE* place to be. You’re relaxed, your brain moving freely, making you prone to an attack by creativity! This is the best time to form great ideas!
Having too much theta waves isn’t desirable though — it could potentially mean that you have ADHD, depression, impulsivity or inattentiveness.
At the same time, having too little could indicate stress and anxiety or poor emotional awareness.
Alpha waves are the default state of our mind. It is evoked when you are physically and mentally relaxed, and not doing something that requires a lot of focus (eg. meditation).
These are also extremely beneficial; studies have shown that alpha waves have a multitude of health benefits. One example of this is an increase in the amount of useful hormones!
Again, too much of a good thing isn’t not good and so, alpha waves could indicate an inability to focus or being too relaxed.
Too little could indicate anxiety (and disorders like Obsessive-Compulsive Disorder), insomnia and stress.
With alpha waves, we come to the end of relaxation-related waves. Beta waves indicate logical tasks and getting things done. This is the type of wave evoked when it’s time to get serious about stuff.
It is also evoked when you are actively alert, feel tensed or afraid.
Too much beta waves could mean arousal, anxiety and stress, while too little could mean ADHD, depression or poor cognition.
Finally, we have gamma waves.
These come into use when you’re pushing through for that last assignment, speedrunning your way through obstacles with expanded consciousness and amazing alertness.
These are also associated with the formation of ideas, language, memory and learning.
Similar to beta waves, Too much gamma could mean high arousal, anxiety and stress, while too little could mean ADHD, depression or learning disabilities.
While these are the typical, we also have a little something called Event-Related Potential. Learn more in the next section!
When records show specific electrical activity owing to specific stimuli, we call it an Event-Related Potential. It could be for something simple like blinking your eye or grinding your teeth. Some common ERPs include:
- P300: This response is elicited in the process of decision making. Essentially, a spike in brain activity of approximately 300 ms is observed when the target stimulus (which is alternated with standard stimulus to create an ‘odd ball’ paradigm) is presented in front of the subject
For your information, an oddball paradigm is an experimental design used within psychology research in which sequences of repetitive stimuli is (infrequently) interrupted with other, different and deviant stimuli. This can take place in different forms like visual and auditory.
- Steady-State Evoked Potentials (response to repetitive stimuli)
It truly is fascinating how our brain has specific patterns or responses based on what’s happening around us. Brain-computer interfaces are all about manipulating these to control machines. As much as I’d love to dive into that, we have a little bit of cleaning to do first…
Time to clean up the mess!
You might be confused — what mess could possibly exist an EEG recording for the sake of crying out loud?
Well, let me ask you this: when you are focused on grinding your teeth, do you not move your eye? Since our brain has a response to literally everything, we need to filter out the noise.
Other potential sources of noise include EEG equipment and external electrical interference.
We clean these in a step called signal preprocessing.
Now, you could clean up this data in many ways, but one of the most common ways of doing so is via filters.
The filters I’m referring to are not Instagram filters or water filters, rather they are electrical circuits that allow alternating current of some frequencies to pass through more easily than others. Kind of like the electrical version of water filters!
Preprocessing usually involves the use of signal and spectral filtering to maximise the ratio of signal to noise.
As of for what to filter, you have to consider what is the data you are expecting. For instance, eye movements may be necessary for one experiment but not another.
After preprocessing, you can use the clean data for actual signal detection.
Making the Data Make Sense
Now that we have our clean data, we need to make it make sense.
This is done in the step of feature extraction. This includes analysing the data and extracting meaningful informations.
This is done using complex mathematical algorithms that can extract informations hidden within the EEG data.
Let’s group our data; Classification/Decoding!
With the feature vectors we got, a classifier/decoder is trained to learn how to detect brain states. These models are ML algorithms which can be trained to group or classify these features.
This step help us figure out what exactly the subject is performing.
Telepathy — Let’s move the football!
Ah, the most exciting part yet! After the signal has been classified, we pass it through a feature translation algorithm, translating the features to the act intended.
…And now you’ve got it! You can now move the football and show all your friends!
While reading this article, a lot of applications might have popped to your head. This technology is nothing short of revolutionary and it can disrupt everyday life — for the better. Imagine robotic limbs controlled by your brain (for the People of Determination)? Or how about video games controlled by your brain? The possibilities are endless in every sense ✨
If you want to jam out on this or just be friends (pretty please), catch me on my socials — Linkedin | Twitter | Instagram | Youtube. Oh, and while you are at it, stay on the loop with my crazy adventures by subscribing to my personal newsletter! Toodles :)
Oh, and here’s a bit about me:
Hey 👋, I’m Rania, a 14 y/o alum innovator in the Knowledge Society. I’m currently researching alternative protein & Neurotech (like brain-computer interfaces). I’m always ready to learn, grow and inspire. I’d love to connect; reach out to me on any of my social media and let’s be friends!