Electroencephalography (EEG) is an electrophysiological monitoring method used to record the brain activity in brain-computer interface (BCI) systems. It records the electrical activity of the brain, is typically non-invasive with electrodes placed along the scalp, requires relatively simple and inexpensive equipment, and is easier to use than other methods. EEG-based BCI methods provide modest speed and accuracy which is why multichannel systems and proper signal processing methods are used for feature extraction, feature selection and feature classification to discriminate among several mental tasks. This edited book presents state of the art aspects of EEG signal processing methods, with an emphasis on advanced strategies, case studies, clinical practices and applications such as EEG for meditation, auditory selective attention, sleep apnoea; person authentication; handedness detection, Parkinson's disease, motor imagery, smart air travel support and brain signal classification.
Inspec keywords: brain; medical disorders; electrocardiography; biometrics (access control); brain-computer interfaces; sleep; hearing; neural nets; medical signal processing; electroencephalography; feature extraction
Other keywords: brain signal classification; auditory selective attention; EEG signal processing feature extraction; EEG signal detection; electrocardiography; electroencephalogram brain waves signals; OSAS blind source separation; biometric brain dermatoglyphic neural architecture; handedness detection system; source analysis; person authentication; motor imagery EEG BCI applications; sleep apnea; Parkinson's disease feature extraction; smart air travel support system
Subjects: User interfaces; General electrical engineering topics; Bioelectric signals; Audition; Electrical activity in neurophysiological processes; Electrodiagnostics and other electrical measurement techniques; Image recognition; Textbooks; Computer assistance for persons with handicaps; Biology and medical computing; Signal processing and detection; Digital signal processing; General and management topics; Pattern recognition
In this study, the focus is placed on measures of brain activities that are associated with meditation. With the intention that a physiological common ground that is universal across differing meditative activities could be found here. Electroencephalogram (EEG) is the selected method of measurement. EEG measurement produces a time series signal that represents the neural activity of the brain. It holds the advantage of being a painless, noninvasive, relatively quick, and economical procedure. Electrodes are placed on the scalp of the subject.
EEG recordings can noninvasively track with high temporal resolution the brain activity associated with different types of stimulus events. By analyzing changes in the event-related potentials (ERPs) as a function of the direction of attention, one can make inferences about the timing, level of processing, and anatomical location of stimulus selection processes in the brain. The scalp electrical potentials that produce EEG are thought to be generated by the extracellular ionic currents caused by dendritic electrical activity (i.e., excitatory and inhibitory postsynaptic potentials EPSP/IPSP). So, what we see in the human EEG is the synchronous excitatory and/or inhibitory input into a large population of nerve cells. With EEG, it is possible to get a glimpse of neural activity from the whole cortex. This makes EEG a very potent tool to study the interaction between brain areas and different cortical networks. This chapter highlights the use of the EEG in the neural correlates of auditory selective attention studies, where an objective quantification measure of the EEG synchrony is introduced to analyze the N1-P2 wave of auditory late responses (ALRs). Studies were carried out with normal hearing subjects as well as tinnitus patients.
To detect the sleep apnea electroencephalogram (EEG) signal, several feature extraction and optimisation methods were investigated in this chapter. The sleep apnea signals were acquired in this experiment. This chapter researched on the abnormalities in EEG for those who suffered from sleep apnea. The statistical correlation measurement of the EEG brain signals was analysed mainly to identify the abnormalities and specific features of patients who suffered from sleep apnea. The features and characteristics of the EEG signals were measured using the fundamental section of the Hilbert-Huang transform (HHT) decomposition method to breakdown EEG sleep apnea data into finite and smaller components. Based on this research, the fundamental empirical mode decomposition (EMD), bivariate decomposition and white noise-based ensemble EMD (EEMD) to the targeted EEG data method were investigated to obtain instantaneous frequency data. All these three methods were used to analyse the extracted sleep apnea EEG signals and features.
This chapter starts with the introduction to various types of authentication modalities, before discussing on the implementation of electroencephalogram (EEG) signals for person authentication task in more details. In general, the EEG signals are unique but highly uncertain, noisy, and difficult to analyze. Event-related potentials, such as visual-evoked potentials, are commonly used in the person authentication literature work. The occipital area of the brain anatomy shows good response to the visual stimulus. Hence, a set of eight selected EEG channels located at the occipital area were used for model training. Besides, feature extraction methods, i.e., the WPD, Hjorth parameter, coherence, cross-correlation, mutual information, and mean of amplitude have been proven to be good in extracting relevant information from the EEG signals. Nevertheless, different features demonstrate varied performance on distinct subjects. Thus, the Correlation-based Feature Selection method was used to select the significant features subset to enhance the authentication performance. Finally, the Fuzzy-Rough Nearest Neighbor classifier was proposed for authentication model building. The experimental results showed that the proposed solution is able to discriminate imposter from target subjects in the person authentication task.
In this project, an algorithm, together with a testing module, has been developed to classify the handedness of a person. Electroencephalogram (EEG) signals at the homologous occipital region were captured when test subjects were asked to rest or expose to some graphical stimulus. Handedness of the person can be determined from the EEG data captured and further confirmed using a simple game as testing module. EEG signals are obtained from three locations, namely A1, O1 and O2. The signals are processed using wavelet transform to classify the signal into four different frequency bands, alpha, beta, delta, and theta, before it is used to find out the mean EEG coherence (MEC). Generally left-handed person has higher MEC, which means that there are more connections between the left and right hemisphere of cerebrums through the corpus callosum (CC).
Parkinson's disease (PD) is a growing disease in Malaysia. This disease affects mostly elderly people aged around 60 years and above. Men are one and half times more likely to have PD than women. Many do not realise until they are diagnosed with PD. An estimated 7-10 million people worldwide are living with PD. The loss of dopamine in substantia nigra (SN) in the brain is the main cause of PD. Insomnia and Bradykinesia are the most common symptoms for every PD patient. These symptoms affect their quality of life. A total of 45 participants took part in an experiment where 21 of them were healthy participants and 24 of them were PD patients. Out of these 24 PD patients, 15 suffered from Bradykinesia and 9 experienced insomnia. Electroencephalogram (EEG) was used extensively to collect data. The data were then processed and evaluated using wavelet to segregate PD patients from healthy adults. The expected outcome for this project is mainly to have a better recognition of the signals collected from PD patients and to identify the similar features in the signals.
The presented analysis scenario shows how source reconstruction can be embedded in the EEG classification system and highlights its benefits for brain-computer interfaces (BCI) performance. Results of the analysis, trained classifiers, and selected feature indices can be directly used in BCI feedback training sessions. Linear inverse operators, such as WMNE, and sparse regions-of-interest are computationally simple enough to be applied in real-time settings.
In this chapter, four experiments were reported. The first study is to identify factors of seating comfort. Next, the survey of relationship between seat location and sitting posture is reported. It is followed by the validation of aircraft cabin simulator. Lastly, the validation experiment of smart neck support system is described. The validation experiment is to validate the developed smart systems for neck support in a simulated “real life”setting. The aim of a developed smart system is to support the passenger's head and reduce passenger neck muscle stress during air travel adaptively. Experimentally measurements were obtained using attached EMG electrodes.
This chapter focuses on brain computer interface (BCI) brain signal classification. BCI classification is a multistep process which includes: brain signal acquisition: This refers to the brain imaging method used to acquire the brain signal, such as electroencephalography (EEG). Preprocessing during the preprocessing step, various signal processing methods such as digital filtering and artefact removal methods are applied in order to improve signal quality. Feature extraction: during this step useful features in the signal associated with the user's cognitive state are extracted. Classification: This involves the extracted features to make predictions about the user's current cognitive state. This can involve machine-learning techniques or other detection algorithms. Device control: this step, commonly known as `translation', involves converting the classifier outputs into a form usable by the external device.
The fingerprints and the brain are connected through the epidermal growth factor (EGF) and neural growth factor (NGF). It is important to know that the fingerprint patterns are linked to the brain's cortical folds.Fingerprints have been proven to be a revelation of the psychology of a person and the understanding of the functioning and behaviour of the brain. Researchers have made tremendous progress with the information gained through the genotypic make up of an individual. The fingerprints are an index of a person's uniqueness based on the study of epidermal ridges on the skin on the fingers and toes. There appears to be a connect between fingerprints and the sulci or cortical grooves in the brain hemispheres.
There are few steps in analysis of ECG signals. The steps are namely noise elimination, cardiac cycle detection, extraction of features from ECG points, formulation of characteristic feature set, and finally classification of the ECG. Currently, we are using machine learning techniques only. The interpretation of the ECG data can be done more efficiently, accurately, and fast using deep learning techniques.
In this research, the electroencephalogram (EEG) signals for sleep apnea were extracted and processed using Blind Source Separation approach. To identify the EEG features, 13 Independent Component Analysis methods were adopted to analyse the data extraction performance. All the EEG signals on Obstructive Sleep Apnea Syndrome (OSAS) were recorded using 10-20 international electrode placement system. The experiment was conducted based on a 20-min sleep recording during rapid eye movement sleep characterized by rapid saccadic movements of the eyes and non-rapid eye movement sleep. Seven electrode positions were identified to record the EEG signals, with a sampling time of 100 Hz. The result was investigated using the proposed 13 ICA algorithms to understand the important EEG signals and features for every process. The wavelets denoising results were obtained to evaluate the robustness of the proposed wavelets denoising algorithms. According to the performance analysis, the proposed wavelets denoising technique could be used to investigate the recorded EEG signals with lower signal amplitude.
