Brain computer interface: From Definition to Working
Brain is considered to be the fastest supercomputer in the world. Since networking of computers is a reality, why can’t we think about a brain-computer network? The story of brain-computer interface begins here. In recent years, the developments in this zone are potentially higher compared to other similar kinds. Many different kinds of applications from various domains are developed and implemented in a rational way. Some major sectors that use brain-computer interface are bioengineering and neuroscience.
Brain-computer interface is a staple of science fiction writing .You may be remembering the aliens from other planets in science fictions communicating through brain-to-brain signal transmission. We are now living in a world where those fictitious things became truth. The power of modern computers helped us to make all such spectacular science fictions into reality. We can create, operate machinery by simple thoughts. Communication became highly personnel among two individuals and done more fast. Don’t think, it’s all about convenience; it’s helpful for unhealthy, severely disabled people around us.
A brain-computer interface (BCI), sometimes called a neural-control interface (NCI), mind-machine interface (MMI), direct neural interface (DNI), or brain-machine interface (BMI), is a direct communication pathway between an enhanced or wired brain and an external device [wiki]. In this definition, the word brain means the brain or nervous system and the word computer means any processing or computational device, from simple circuits to silicon chips. Early researches on BCIs began in 1970s at the University of California. Traditional BCIs allow a user to communicate or control a computer by brain activity .Recently they are used to extract information about the user and infer his mental states (e.g. workload, vigilance).
Different Devices used in BCI
Brain activity can be monitored using different devices such as:
- scalp-recording Electroencephalogram (EEG)
- Magnetoencephalogram (MEG)
- Functional magnetic resonance imaging (fMRI)
- Electrocorticogram (ECoG)
- Positron emission tomography (PET)
- Near infrared spectroscopy (NIRS)
Among these fMRI, NIRS, PET depends any type of brain imaging technology which rely on changes in blood flow (hemodynamic response) and MEG, EEG measures the brain’s magnetic and electrical activity respectively. Conventional BCIs have lack of high accuracy, reliability, low information transfer rate, and user acceptability. A new approach to create a more reliable BCI that takes advantage of each system is to combine two or more BCI systems with different brain activity patterns or different input signal sources. This type of BCI is called Hybrid BCI. It reduces disadvantages of each conventional BCI system [1].
Types of BCI
We can divide the BCI’s into three categories as Non-invasive, Semi-invasive and Invasive depending on the method used to collect brain signals.
Non-invasive method
It measures the brain signals by placing the sensors (electrodes) on the scalp, the most external part of the brain. There exist several non-invasive techniques used to study the brain. EEG is the most common and cost effective way. Some other technologies applying this non-invasive method are MEG, PET , fMRI, fNIRS.
Semi-invasive method
In this method the electrodes are placed inside the skull but outside the brain, on the exposed surface of the brain. In terms of brain science, the electrodes are placed either outside the dura matter (epidural) or under the dura matter (subdural). Large number of electrodes is present in the strip or grid of electrodes allowing us to cover large areas of the brain. ECoG is the leading technique used in this method.
The spatial resolution of ECoG is much higher because the signal doesn’t have to travel to reach the scalp. The spatial resolution in ECoG is tenths of millimeters, while it is centimeters in EEG [3]. ECoG electrodes are placed on the brain only during necessary medical surgery.
Invasive method
Invasive types of BCI’s are implanted directly into the grey matter of the brain during neurosurgery. The micro-electrodes are placed directly into the cortex, measuring the activity of each single neuron. There are single unit BCIs, which detect the signal from a single area of brain cells, and multiunit BCIs which detect from multiple areas. Electrodes have different lengths, for example, up to 1.5 mm (Utah, Blackrock Microsystems) or 10 mm (FMA, MicroProbes) in a micro-electrode array (MEA) [2]. The quality of the signal is the highest, but it includes several problems, as for example forming scar tissues. The body reacts to the foreign object and builds the scar around the electrodes, which cause deterioration in the signal. Due to this risky and expensive process, the target of invasive BCI is mainly blind and paralyzed patients.
How BCI works
The working of any BCI can be described in three main steps: collecting brain signals, processing and interpreting signals, finally outputting commands to a connected machine through an application interface. More BCI systems are operating in non-invasive method by collecting signals from the scalp of the brain. The central element in each BCI is a translation algorithm that converts electrophysiological input from the user into output that controls external devices. BCI operation depends on effective interaction between two adaptive controllers: the user who encodes his or her commands in the electrophysiological input provided to the BCI, and the computer which recognizes the command contained in the input and expresses them in the device control [4].
The common mechanism of the brain computer interface can be summarized as follows.
Step 1: Signal Acquisition
The brain signals are captured from the brain during events/tasks through invasive or non-invasive methods. These signals are stored and digitized without signal loss. In case of weak signals, amplification of signals has done using hardware equipment.
Step 2: Signal Pre-Processing
Most of the time, the incoming data will be contaminated due to environmental and physiological factors. It becomes necessary to carry out different pre-processing techniques. Pre-processing steps are followed not only for artifact removal, but necessary for normalization and standardization of data.
Step 3: Signal Classification
Once the pre-processing is over and signals are fairly clean, they will be processed and classified using certain machine learning algorithms to find out which kind of mental task the subject is performing.
Step 4: Device Interaction
The last step in BCI applications is the building up of an appropriate algorithm for the development of a certain application. It makes use of the decoded information from brain signals through the classification algorithm and interacts with the controlling device/application in a wired or wireless mode to operate accordingly.
Brain computer interface applications are still in the primary level and have a lot of limitations. We are unable to decode the neuronal activity in a full-fledged manner. The pathway of neuron transmissions are still an area of research. Most of the signal acquisition methods are non-invasive and these weak electric signals from brain are prone to interference. At the signal acquisition step, signal-to-noise ratio (SNR) and signal-to-interference (S/I) ratio are the key attributes that depends on the accuracy of the system. Some of the devices have good spatial resolution and some other have better temporal resolution. For an efficient BCI system, we need optimal temporal and spatial resolution to signals.
BCI applications
Earlier BCI research and development works focused mainly on neuroprosthetics which help to restore damaged sight, hearing and movement. After many experimentations, in June 2004, Matthew Nagle became the first human to be implanted with BCI, Cyberkinetic’s BrainGateTM. Another remarkable development is made by Jonathan Wolpaw et.al in December 2004, they came up and demonstrated the ability to control a computer using a BCI. In this study, subjects are asked to wear an electrode cap to capture EEG signals from the motor cortex-part of the cerebrum governing movement [5]. BCIs had a long history of control applications like cursor controls, motor-movement control, robotic arms, telephone dialing etc. A wide range of applications are now upcoming in BCI facilitating various technologies. Some of the prominent BCI application areas are given below.
1. Medical applications
2. Robotics
3. Gaming and entertainment
4. Educational and self-regulation
5. Security and authentication
6. Neuromarketing and advertisement
7. Neuroergonomics and smart environment
REFERENCES
1. Amiri, S., Fazel-Rezai, R., & Asadpour, V. (2013). A review of hybrid brain-computer interface systems. Advances in Human-Computer Interaction, 2013.
2. Waldert, S. 2016. Invasive vs. Non-Invasive Neuronal Signals for Brain-Machine Interfaces: Will One Prevail? Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4921501/
3. Leuthardt, E. C. et al. 2004. A brain–computer interface using electrocorticographic signals in humans. Journal of Neural Engineering, Volume 1, Number 2. Available from: http://iopscience.iop.org/article/10.1088/1741-2560/1/2/001/meta;jsessionid=E687A7B4A1215A8655C2DF6429F87A05.c2.iopscience.cld.iop.org
4. Wolpaw, J. R., Birbaumer, N., Heetderks, W. J., McFarland, D. J., Peckham, P. H., Schalk, G., … & Vaughan, T. M. (2000). Brain-computer interface technology: a review of the first international meeting. IEEE transactions on rehabilitation engineering, 8(2), 164–173.
5. Available at website :https://towardsdatascience.com/a-beginners-guide-to-brain-computer-interface-and-convolutional-neural-networks-9f35bd4af948
