Understanding What Are Brain Computer Interfaces Psychology Essay
A Brain-Computer Interface can be defined as a communication system which enables a user to communicate without making any muscle movement. Our brain generates signals which can be analyzed to decode user’s intention. EEG (Electroencephalography) is the most widely used methodology for measuring the ongoing brain activity. Most of the BCIs utilize the neurological phenomenon known as steady state visual evoked potential (SSVEP). Visual evoked potential results when a subject is presented a visual stimulus, visual stimuli can be anything like a flickering checkerboard or a light flash. Evoked potentials are highly corrupted by noise, the signal to noise ratio can be as low as 1:100, and the noise is eliminated by employing some signal processing techniques. A user, with the help of these SSVEPs, can command a robot to move in any direction. The neurological principle utilized is that when a visual stimulus is applied, ranging from 3.5 to 100Hz, the brain also starts generating signals with the same frequency. Accurately determining these closely spaced frequencies enables to distinguish between user’s intensions. An effort is made to look at various signal processing algorithms for de-noising the evoked responses of the brain and comparison between high resolution frequency estimation techniques is made for employing the best one to enhance the performance of the BCIs.
Brain Computer Interfaces (BCIs) have gained a lot of interest over the last decade. A BCI is basically a communication channel between a user and a machine. BCI finds a lot of applications for physically disabled people as the output of BCI don’t require any muscular movement. The inputs to a BCI are the brain signals generated and based on these brain signals the corresponding output is produced.
Depending on the method employed for measuring the brain signals, BCIs are classified as invasive and non-invasive BCIs [5, 6]. Invasive methods are harmful for the user, so mostly the non-invasive methods are employed like Electroencephalogram (EEG) or Magnetic Resonance Imaging (MRI). EEG is the most widely used method for effectively measuring the ongoing brain activity. A basic block diagram of a BCI can be presented as:
C:\Users\engguest\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.IE5\A92R3YBR\MC910216349.png
Figure 1: Basic Block Diagram of a Brain Computer Interface (BCI)
The Electroencephalogram (EEG) signals are recorded with international 10-20 system of electrodes which are mounted on a human scalp. There are two types of signal patterns that can be observed with EEG signals namely: Evoked Potentials (EPs) or Event Related Potentials (ERPs) and Event Related Changes in present EEG (ECoG). Evoked Potentials occur when a certain stimuli is presented to a subject. ECoG on the other hand are observed in a particular range of frequencies and as the name itself suggests they are the changes in the current EEG signals.
The other important part of a BCI is the way in which the brain activity is modified; this process is called as ‘Mental Strategy’ . A person may be asked to concentrate on certain things like taking a note of the changes in the ongoing visual pattern or to imagine moving his/her limbs. The signals generated by brain can then be utilized for designing an appropriate system.
There are basically two modes of operation for a BCI; synchronous and asynchronous respectively . For a synchronous BCI a fixed and beforehand defined time frame is utilized. A user has to develop a particular mental state whenever a cue-stimulus is presented. For an asynchronous mode BCI the EEG signals are continuously monitored and the system has to itself take a note of the points when a particular mental state occurred. Asynchronous BCIs are more complicated as compared with a synchronous one but they offer a large amount of flexibility as the user doesn’t have to worry about acting to produce a definite state of mind.
A BCI is a closed loop system. It has a feedback and it is very important. Along with the feedback a BCI also has two adaptive controllers namely brain and the computer. A feedback can be continuous or discrete depending on the application.
The most important part of a BCI is the feature extraction unit. The message or command which is passed on by the user is encoded in the brain signals. The feature extraction unit has to clean the signal received that is the noise is removed and then the desired signal or command sent by the user is decoded from the noise free brain signals.
History of Brain Computer Interfaces:
The work on Brain Computer Interfaces started long time back. In 1970s the Advanced Research Project Agency (ARPA), a government organization of the United States decided to go ahead with the research in this field. Their objective was to increase the efficiency of heavy mental tasks by improving the man’s capabilities with machine power.
In eighties Wolpaw developed a BCI that controlled a cursor movement using the frequencies that centered around 9 Hz. This system is also called as Albany BCI and it used autoregressive parameters .
Another system known as Thought Translation Device was developed by Birbaumer’s team. It used the self regulated slow cortical potential shifts . After going through a lot of training the users can efficiently write the texts.
Graz BCI team developed a system which used the thoughts in order to make a person walk in a virtual street. This system was practically tested in the year 2004 in the Computer Animated Virtual Environment at University College of London .
Figure 2: Hand Grasp for a person suffering from heavy spinal cord injury .
Donchin et al. another research team developed a BCI which recognized the character thought or picked by a user. A letter matrix of size 6X6 was presented to the user which had a row or a column flashing. The visual evoked potential utilized for this BCI is P300.
Figure 3: Walking in a virtual street by ‘thought’.
Flickering of a light source can also be utilized for developing a BCI. The BCIs which works on a light source i.e. a flickering checkerboard or a flashing LED, are called as BCIs based on Steady State Visual Evoked Potentials (SSVEPs). The neurological phenomenon utilized here is that when a person concentrates on a light source flickering at a particular frequency (ranging between 3.5 to 100 Hz) the brain also starts generating electrical signals with the same frequencies. This phenomenon has been very effectively utilized for BCIs and with these BCIs a data transfer rate of up to 90 bits/minute can be achieved .
2.2 Applications of Brain Computer Interfaces:
Many movies show human beings controlling big machines. These things look fascinating but at the same time they don’t appear to be real. Now let’s think wont it be nice if a person controls a huge machine just by the power of his thoughts? If this question was asked a decade ago it would have definitely sounded impossible. But, now a day’s Brain Computer Interface provides us with the power of controlling things just by our thoughts.
A BCI find its application almost everywhere. It can be utilized by handicaps, paralyzed patients or by the defense services. For people who have been suffering from major spinal cord injuries or those who cannot make their limbs move, BCI provides a great helping hand. Spelling systems have been developed for physically disabled people. Neuroprosthesis is another field which is utilizing BCIs a lot. BCIs are being employed to move the prosthetic limbs of a patient or making another movements like a hand grasp.
A lot of BCIs have been developed where a tetra-plegic patient can make a complex movement like grasping a can with his hand. These BCIs take the dynamics of the brain into consideration. The working of the brain is closely observed and the BCI makes the necessary movement when the person wants to grasp a can.
Another BCI can control the movement of a person in virtual street. A person can walk in the forward direction by imagining the movement of his/her legs. These intensions are sensed by the BCI system and corresponding movement is made.
Lately it has been experimentally proved that monkeys can do a 3-D tracking. They think about the correct path and the BCI makes the move in the same direction. Sooner or later this could well be implemented on human beings as well.
In the coming future the signal recording will be the field that’s going to gain a lot of attention. The methods which are currently employed for the brain signal measurements are all non invasive techniques i.e. the sensors are placed on the scalp not under the scalp. But in near future invasive methods of signal recording will come into picture.
An Example of a BCI based on Steady State Visual Evoked Potential:
Presently there are various BCIs in the market which employs the Steady State Visual Evoked Potentials. The figure 4 below displays one of such BCI. Where the system consists of (1) a steady-state visual evoked potential based brain computer interface (BCI), (2) a laptop controlled iRobot platform. The laptop on the robot platform sends robot's view videos to the operator/subject via Skype Video Chat and receives navigation commands from the operator/BCI via remote connection. The BCI operates on the premise that the visual cortex synchronizes with the periodical checkerboard flickering when the subject directs his gaze at one checkerboard and a spectrum estimator based frequency detector determines which of the 4 checkerboards is selected. Each checkerboard flickers at 7, 9, 11, 13Hz respectively and they correspond to left/right turn, go forward, and stop commands for the robot. The BCI also features a "reject" class for the instances that the operator is looking at the video and not a checkerboard. The robot executes the last received command until a new one is received .The system is able to achieve 100% accuracy with a delay of less than 1 second for the subject in the picture.
Figure 4 : BCI where four checkerboards flashing at frequencies 7, 9, 11, and 13 Hz respectively are presented to the subject, the subject has an electrode cap mounted on the scalp for recording the EEG activity .
CHAPTER III. EVENT RELATED POTENTIALS (ERPS) OR EVOKED POTENTIALS (EVPS)
The brain generates electrical signals which can be easily recorded with Electroencephalography (EEG). Whenever a human being is applied any stimulus a corresponding change in the ongoing pattern of the EEG is observed. This change is known as the brain response to the stimuli technically called as Evoked Potentials (EPs) or Event Related Potentials (ERPs).
Every action we take is controlled by the electrical signals which run through our body. We already know that there is nothing which isn’t made of atoms. Atoms comprises of neutrons (neutral), protons (positive charge) and electrons (negative charge). When there is an imbalance in the charge i.e. when the positive charge of an atom is not equal to its negative charge the atom acquires either positive or negative charge (depending on whichever is greater in quantity protons or electrons). This change in the charge between two atoms allows the electrons to flow and hence giving rise to electricity. As human body is also a big mass of atoms, it also gives rise to electricity. When a brain conveys signal to a particular body part, like the movement of the limbs, it does that by transporting the charge. The cells work as charge carriers and the charge is carried till it reaches the target.
There are a large number of neurons present in the brain. The charge inside the brain is maintained by the neurons. For transporting the charge the ions are pushed across the membranes. Whenever a neuron receives any signal from its neighbor it discharges the ions in the region which lies out of the cell. Ions having the same charge repel each other they keep on pushing their neighbors which results in a wave. This procedure is in general called as ‘volume conduction’. Now, whenever this wave reaches the scalp (where there are electrodes placed), due to the charge present the electrons, present at the electrodes, are either repelled or attracted. This difference of push or pull or in other words difference in the voltage between two electrodes can be recorded by using a voltmeter. Later the process of EEG is discussed in this report .
3.1 Classification of Evoked Potentials:
Corresponding to the stimuli applied the Evoked potentials are classified as Auditory Evoked Potentials (AEPs), Visual Evoked Potentials (VEPs) and Somatosensory Evoked Potentials (SEPs). Auditory Evoked Potentials results when a human is presented stimuli in the form of a click similarly a visual stimuli can be a flickering checkerboard or a light flash, Somatosensory Evoked Potential is produced when a person touches a hot object.
3.1.1 Auditory Evoked Potentials (AEPs):
As mentioned earlier AEPs results whenever an auditory stimuli is presented to a human being. Auditory stimuli can be a click or a tone, after applying these stimuli a change in the ongoing pattern of the EEG is seen which is called as Auditory Evoked Potentials. Depending upon the response time AEPs are further classified into early, middle and late latency Auditory Evoked Potentials.
Early AEPs are a combination of a couple of responses which results in the first 12 ms after the stimuli is applied. The reaction which is recorded in the first 2.5 ms after applying the stimuli arises from the cochlea as well as the auditory nerve. The brain stem also produces the reaction in the first 12 ms; these reactions are captured from the vertex and are called as Brain Stem Auditory Evoked Potentials (BSAEPs). The early AEPs are useful in detecting problems with the functioning of auditory pathway of a human being. Taking the example of small children who cannot speak or other disabled people who have difficulty in communicating with others, early AEPs serve as the best way of knowing if there is anything wrong with the normal auditory pathway. When a person recovers from coma then early EVPs can be very useful in detecting the patient’s condition.
Middle AEPs occur between 12 ms to 50 ms, they are just like sinusoid wave having positive and negative peaks. These peaks are noticed during the time frame of 12 ms to 50 ms after the stimulus is applied to the subject. They are not so useful for clinical applications as early AEPs because there is still a lot of doubt as far as their locations are concerned.
Late AEPs are noticed between 50 ms to 250 ms after the stimulus is applied. They basically comprises of four main lobes or peaks called as P50, N100, P150 and N200, the meaning of these peaks will be discussed later in this report. These peaks have maximum amplitude when they are recorded from the vertex and they arise from the cortex.
3.1.2 Somatosensory Evoked Potentials (SEPs):
SEPs results when a stimulus is applied to any of the peripheral nerves. They are usually employed to check whether everything is fine with the somatosensory pathway. Diseases like multiple sclerosis can be diagnosed using the SEPs. During the surgeries a person is anesthetized so it becomes very essential to keep a track of any kind of a neurological damage, SEPs play a very vital role during surgeries. Keeping a track of the functioning of spinal cord is done using SEPs.
If you are the original writer of this essay and no longer wish to have the essay published on the UK Essays website then please click on the link below to request removal: