In order to understand what EEG is and it’s working, we need to understand basic anatomy and functional areas of the brain. This article will also provide you a basic overview of that.
Our brain controls everything, be it unconscious tasks like breathing or consciously made decisions like learning a new skill, every decision is controlled by our brain. The building block of our brain are nerve cells called neurons(shown in figure below). These nerve cells transmit information throughout the body in the form of electrical impulses.
Neurons comprises of three major sections Dendrites, Cell body(Soma) and Axon as seen in the figure below. Neurons are connected to other neurons in a complex yet well defined circuitry which still is not completely understood by modern science. The dendrites are connected to other neurons and acts as a receiver, they take the information to the cell body. Then axons carry the electrical current to the terminals where they transmit the information forward.
This network of neurons act as a complex network of computers forming the internet. A simplified example could be, if you wish to watch a YouTube video on your computer, the cable from your internet provider which comes to your house is the dendrite. Your router acts as cell body which decides on where to route the video and finally the ethernet cable is the axon which carries the video to your computer.
The outer most layer of nerve tissue in an human brain is know as cerebral cortex, and it is naturally separated in four lobes namely frontal lobe, parietal lobe, temporal lobe, and occipital lobe(see image below). Each of these lobes has a discreate set of functions as well as connections to the other part of the brain.
The frontal lobe is responsible for immediate and sustained attention, social skills, emotions, empathy, time management, working memory, and character. The frontal lobe is generally known as an executive planner i.e. it helps in maintaining control, plan for the future, and monitoring the behavior.
The parietal lobe works as a navigation system for our body, it integrates all the information from different senses and generates a coherent representation of the environment. For example, assume you are reading a book and having a cup of tea. Now every time you don’t have to look where exactly your cup is, while focusing on the book you can pick up the cup have your tea and then put it back. This is possible because the parietal lobe stores and retrieves the shape, size and orientation of objects to be grasped.
The temporal lobe houses auditory cortex which process the auditory signals and it is also involved in memory-making process especially verbal memory. The Left temporal lobe is involved with language interpretation written and verbal whereas the right temporal lobe is more involved when listening to music, and understanding social cues.
The occipital lobe is where all the visual processing takes place. Anything in our visual field seen through our eyes are routed to the occipital lobe where it is processed and hence it has strong connection to the entire brain networks be it the frontal lobes for problem definition, parietal lobe in locating objects or memory regions which are present under the cerebral cortex.
As seen above, different regions of the brain are constantly communicating with each other. This communication is accomplished via neurons. When clusters of neurons fire in synchrony using electrical impulses they produces an electric potential which is captured by the sensors places on the scalp.
Since these signals are originated deep within the brain, voltage captured by the sensors is in μVs(1/1,000,000th of a Volt), usually between 10-50 μVs. This is extremely low and can be easily corrupted very easily with small eye movements, muscle movements, sweat etc. This particularly makes it very difficult processing EEG data and hence the biggest limitation of EEG devices is that the user has to be sitting calm and relaxed to capture any usable information.
Capturing EEG signal which could be used for analysis later is a bit complex then just placing a sensor over the head. 3 sensors are used to capture 1 channel: two active and one ground. This is done using differential amplifiers, which sounds scary but all it does is it returns the difference between two active sensor values. Have a look at the image below the red and the green sensors are the active sensors represented by V1 and V2. The output is the difference of voltages captured at those location. Yellow sensor is referred as the ground sensor, which is usually placed on the ear lobe or mastoid bones to minimize the activity captured by these sensors.
The International Federation of Societies for Electroencephalography and Clinical Neurophysiology, introduced the conventional electrode placement, also known as the 10-20 system. Like any other scientific system, it was introduced to maintain a standardized system to ensure that any study can be reproduced and effectively analyzed.
The international 10-20 system contains 19 sensor locations, and each sensor location is designated based on the lobe it is placed on. Odd numbers are on the left, even numbers on the right and Z on the center line.
In EEG signals, it is observed that they have certain characteristics. These characteristics change as a person ages as well as the state the person is in (sleeping or awake). Due to these characteristics, brain waves can be broken up into five categories. These 5 categories are called alpha (α), theta (θ), beta (β), delta (δ), and gamma (γ) and represent a band of frequencies. Image below shows the frequency range of each category as well as some of the mental functions associated to those categories.
The following image is a snapshot from an actual EEG recording, where the brown lines are 1 second marks. If you observe closely, a combination of alpha(red), beta(yellow) and gamma(grey) bands can be observed in the data, which was expected because during the recording the user was sitting with eyes closed but in a wakeful state.
 Demos, J.N., 2019. Getting Started with EEG Neurofeedback. WW Norton & Company
 Burger, C., 2014. A novel method of improving EEG signals for BCI classification (Doctoral dissertation, Stellenbosch: Stellenbosch University).