A system that converts a user's brain activity patterns into messages or commands for an interactive application is known as a BCI. The term "brain-computer interface" (BCI) refers to a control and/or communication system where the user's orders and messages are independent of standard brain-motor periphery communication pathways. EEG is frequently used to assess the brain activity of BCI users. Recording neural activity, extracting features, obtaining crucial information, and integrating that information for beneficial reasons are the four main processes that BCI typically uses to operate. The four steps of the EEG signal analysis include the collection of raw EEG data, signal pre-processing, feature extraction, and classification.
The steps of signal analysis, researchers summarized like this. All of the EEG data that was gathered without any pre-processing or additional analysis is included in the raw EEG data collection. Pre-processing is used to remove noise and improve the quality of raw EEG data gathered for further analysis after it has been recorded. Discriminative and non-redundant information from the EEG data is extracted during feature extraction to create a set of features. The various signal properties that are captured by extracted features can be used as a foundation for differentiating between task-specific brain states. The task performed by a person is identified during the classification step, and the system responds appropriately. Utilization of a particular feature extraction and categorization technique is dependent upon specific application of BCI. (Dr. Padfield)
BCI are typically categorized according to how intrusive they are. Depending on whether the electrodes are inserted into the brain and the device is physically attached to the brain, or if it reads brainwaves from the top of the head, the BCI may be invasive or non-invasive. Direct electrode contact with the brain using an invasive method is more effective because there are fewer interfering variables that affect the signal quality, but this method has risks connected to surgery and its repercussions. The non-invasive approach, which reads brainwaves from the top of the scalp using electrodes on the scalp, is more common and simple. There is a lot of ongoing research to find the best methods for signal recognition and analysis because this method also has certain limitations, such as interference from outside noise, influences from the subject's posture and mood, and recognizing a low strength signal resulting in a reduced signal quality.
EEG detection is now the most used non-invasive technique for obtaining brainwaves for BCI. The low cost of the equipment, fewer risks compared to invasive treatments, portability, simplicity of use, and capacity to monitor cerebral activity directly all contribute to the popularity of EEG. EEG reading is simple to use and has a high likelihood of being put to use by the majority of people in order to take benefits of the opportunities that BCI provides. Magnetoencephalography (MEG), near-infrared spectroscopy (NIRS), and functional magnetic resonance imaging (fMRI) are further non-invasive techniques that can be employed singly or in combination. High temporal resolution, exceptional portability, low cost, less invasive compared to fMRI, and independence from external devices are all benefits of EEG as compared to MEG. (Dr. Kaido)
Depending on how the brain signal is detected, transmitted to the BCI application, and how the BCI itself operates, different subgroups of BCI could be created. Depending on how the BCI application is controlled, BCIs can be classified as either active or passive. The classification of BCIs according to active or passive control, as well as the associated methods for gathering EEG data, are as follows.
1. Motor-imagery: envisioning the motion of a particular body part, such as the hands, feet, or tongue. The EEG could reveal the purpose by observing how it affects brain activity. The parts of the brain that produce real movement are stimulated by imagining.
2. Visual evoked potential: External visual stimulus has an impact on brain activity, and the accompanying changed EEG activity is noted. For instance, steady-state visual evoked potential (SSVEP) involves a variety of flickering visual stimuli that occur at various frequencies, and depending on the subject's gaze direction, the EEG pattern will be congruent with the particular flashing rate.
3. Auditory evoked potential: Depending on the subject's concentration, auditory stimulation is produced, and associated EEG activity is recorded.
4. Vibrotactile evoked potential: Different physical vibrations are produced, for instance on the subject's hands and feet. A corresponding EEG pattern to the particular physical vibration is registered depending on the subject's focus and might be utilized to operate an external device.
5. Imagined speech: Words or sentences that are identified from the EEG are imagined.
6. Error-related potential: When there is a discrepancy between the subject's intents and the BCI application's response, the error-related potential is created. The method can be applied to fix assignments that the subject has supplied. For instance, if the subject moves the cursor in the wrong direction while controlling it, an error-related potential is generated, and the cursor's path can be changed.
1. Analyzing EEG spectral changes: systems where brain signals produce results without the need for conscious effort. Monitoring things like sleepiness, focus, mental workload, emotions, concentration, and other mental states, for instance.
The BCIs that use active control respond to deliberate attempts to change brainwave patterns, and the applications can be actively controlled by the user. The BCIs that use passive control respond to the brainwaves' involuntary condition, such as that of meditation, excitement, or stress. For instance, prepared tasks, audio-visual elicitation utilizing short cinema video snippets, or visual-based elicitation using images could all be used to elicit different emotions.
Different methods could be used to detect the signal while the BCI application is in active control. The scope of these methods is wide and includes hybrid, error-related potential, motor imagery, external stimulation (such as visual, auditory, and vibrotactile), and other methods. For instance, during the motor imaging task, the subject imagines moving a certain body part, and during the error-related potential task, error-related potential is produced when the subject's intents and the BCI application's response don't match. (Dr. Yar)