This chapter presents a summary of the issues discussed during the one day workshop on “Support Vector Machines (SVM) Theory and Applications” organized as part of the Advanced Course on Artificial Intelligence (ACAI ’99) in Chania, Greece [19]. The goal of the chapter is twofold: to present an overview of the background theory and current understanding of SVM, and to discuss the papers presented as well as the issues that arose during the workshop.

Support vector machines: Theory and applications

The performance of motor imagery (MI) based Brain-computer interfacing (BCI) is easily affected by noise and redundant information that exists in the multi-channel electroencephalogram (EEG). To solve this problem, many temporal and spatial feature based channel selection methods have been proposed. However, temporal and spatial features do not accurately reflect changes in the power of the oscillatory EEG. Thus, spectral features of MI-related EEG signals may be useful for channel selection. Bispectrum analysis is a technique developed for extracting non-linear and non-Gaussian information from non-linear and non-Gaussian signals. The features extracted from bispectrum analysis can provide frequency domain information about the EEG. Therefore, in this study, we propose a bispectrum-based channel selection (BCS) method for MI-based BCI. The proposed method uses the sum of logarithmic amplitudes (SLA) and the first order spectral moment (FOSM) features extracted from bispectrum analysis to select EEG channels without redundant information. Three public BCI competition datasets (BCI competition IV dataset 1, BCI competition III dataset IVa and BCI competition III dataset IIIa) were used to validate the effectiveness of our proposed method. The results indicate that our BCS method outperforms use of all channels (83.8% vs 69.4%, 86.3% vs 82.9% and 77.8% vs 68.2%, respectively). Furthermore, compared to the other state-of-the-art methods, our BCS method also can achieve significantly better classification accuracies for MI-based BCI (Wilcoxon signed test, p < 0.05).

/pdf/bispectrum-based-channel-selection-for-motor-imagery-based-299irdfwhv.pdf

Bispectrum-Based Channel Selection for Motor Imagery Based Brain-Computer Interfacing

BACKGROUND An increase in lifespan in our society is a double-edged sword that entails a growing number of patients with neurocognitive disorders, Alzheimer's disease being the most prevalent. Advances in medical imaging and computational power enable new methods for the early detection of neurocognitive disorders with the goal of preventing or reducing cognitive decline. Computer-aided image analysis and early detection of changes in cognition is a promising approach for patients with mild cognitive impairment, sometimes a prodromal stage of Alzheimer's disease dementia. METHODS We conducted a systematic review following PRISMA guidelines of studies where machine learning was applied to neuroimaging data in order to predict whether patients with mild cognitive impairment might develop Alzheimer's disease dementia or remain stable. After removing duplicates, we screened 452 studies and selected 116 for qualitative analysis. RESULTS Most studies used magnetic resonance image (MRI) and positron emission tomography (PET) data but also magnetoencephalography. The datasets were mainly extracted from the Alzheimer's disease neuroimaging initiative (ADNI) database with some exceptions. Regarding the algorithms used, the most common was support vector machine with a mean accuracy of 75.4%, but convolutional neural networks achieved a higher mean accuracy of 78.5%. Studies combining MRI and PET achieved overall better classification accuracy than studies that only used one neuroimaging technique. In general, the more complex models such as those based on deep learning, combined with multimodal and multidimensional data (neuroimaging, clinical, cognitive, genetic, and behavioral) achieved the best performance. CONCLUSIONS Although the performance of the different methods still has room for improvement, the results are promising and this methodology has a great potential as a support tool for clinicians and healthcare professionals.

/pdf/machine-learning-methods-for-predicting-progression-from-4ouw9072j3.pdf

Machine learning methods for predicting progression from mild cognitive impairment to Alzheimer’s disease dementia: a systematic review

Grey relational analysis method based on cumulative prospect theory for intuitionistic fuzzy multi-attribute group decision making

BACKGROUND Commercially-made low-cost electroencephalography (EEG) devices have become increasingly available over the last decade. One of these devices, Emotiv EPOC, is currently used in a wide variety of settings, including brain-computer interface (BCI) and cognitive neuroscience research. PURPOSE The aim of this study was to chart peer-reviewed reports of Emotiv EPOC projects to provide an informed summary on the use of this device for scientific purposes. METHODS We followed a five-stage methodological framework for a scoping review that included a systematic search using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR) guidelines. We searched the following electronic databases: PsychINFO, MEDLINE, Embase, Web of Science, and IEEE Xplore. We charted study data according to application (BCI, clinical, signal processing, experimental research, and validation) and location of use (as indexed by the first author’s address). RESULTS We identified 382 relevant studies. The top five publishing countries were the United States (n = 35), India (n = 25), China (n = 20), Poland (n = 17), and Pakistan (n = 17). The top five publishing cities were Islamabad (n = 11), Singapore (n = 10), Cairo, Sydney, and Bandung (n = 7 each). Most of these studies used Emotiv EPOC for BCI purposes (n = 277), followed by experimental research (n = 51). Thirty-one studies were aimed at validating EPOC as an EEG device and a handful of studies used EPOC for improving EEG signal processing (n = 12) or for clinical purposes (n = 11). CONCLUSIONS In its first 10 years, Emotiv EPOC has been used around the world in diverse applications, from control of robotic limbs and wheelchairs to user authentication in security systems to identification of emotional states. Given the widespread use and breadth of applications, it is clear that researchers are embracing this technology.

/pdf/10-years-of-epoc-a-scoping-review-of-emotiv-s-portable-eeg-21bc7wcox9.pdf

10 years of EPOC: A scoping review of Emotiv’s portable EEG device

Brain-computer interface (BCI) systems translate the human neurophysiological activities into commands through EEG analysis. Improving the BCI performances leads to faster and easier use and less fatigue. In this study, we proposed a new prepossessing approach to increase the robustness of a steady-state visual evoked potential (SSVEP) based BCI. Inspiring from the known properties of the SSVEP frequency components, the goal was to enhance the signal quality by making it more convenient to be interpreted by the decision-making step. We first investigated the potential to detect the deteriorating periods based on the physiological properties of the SSVEP. The proposed system localizes the intervals which can obscure the SSVEP frequencies by a new algorithm founded on the processing and the analysis of the instantaneous phase. The piecewise linear regression allows a sampler comprehension of the phase signal. Then, these intervals are filtered by the moving average filter to enhance the SSVEP quality. Finally, the decision making is made by the canonical correlation analysis (CCA) algorithm. The results of experiments, using real EEG signals from five subjects, show that the proposed approach significantly increases the performances in terms of accuracy and information transfer rate by about 7.3% and 3.85 bits/min, respectively, in case of 2 s segment length. On the other hand, the spatial filtering methods of the literature weaken the system performances.

/pdf/ssvep-enhancement-using-moving-average-filter-controlled-by-2ssg7p6ra2.pdf

SSVEP Enhancement Using Moving Average Filter Controlled by Phase Features

Brain-Computer Interface (BCI) systems are widely based on steady-state visual evoked potentials (SSVEP) detection using electroencephalography (EEG) signals. SSVEP-based BCIs are becoming attractive due to their higher signal-to-noise ratio (SNR) as well as faster information transfer rate (ITR). However, their performances are largely affected by the interference coming from the spontaneous EEG activities which intrinsically restrict their efficiency in distinguishing between SSVEPs and background EEG activities. In this paper, we introduce a new approach for the detection of SSVEP based on bispectral analysis to palliate the frequency-dependent bias. A COMB filter associated with a wavelet denoising filter is firstly used to minimize the noise while improving the SNR of phase signals. Next, the complementary orthogonal projections and the principle component analysis (PCA) are used to decompose the components related to SSVEPs and components related to brain activities. Finally, the bispectrum, a powerful tool for the analysis and the characterization of nonlinear properties of stochastic signals, is used to extract the features of the EEG signal benefiting from the information about the phase coupling of the signal components. The results of experiments, using two databases on five (or ten) subjects, show that the proposed approach significantly outperformed the standard CCA approach in distinguishing the target frequency and in average information transfer rate.

Bispectral analysis-based approach for steady-state visual evoked potentials detection

Recently, the low cost EEG acquisition systems such as the Emotiv Epoc give new tools to develop Brain-Computer interface systems for everyday use outside the laboratory. However, the low sampling rate and the low number of channels remain possible sources of failure. The Canonical Correlation Analysis and the Multivariate Synchronization Index (MSI) methods are applied in a SSVEP-based BCI in order to compare their accuracies. The main goal of this research is to find the appropriate method allowing the control of an autonomous wheelchair by the severely handicapped people. The experimental results show that the MSI method reaches 96% of accuracy with optimal parameters.

A comparison study of SSVEP detection methods using the Emotiv Epoc headset

The automatic seizure detection system is designed to aid the physician's decision-making process with recognizing the sought EEG segments. Increasing the system sensitivity is the goal of several studies. In fact, ameliorating this criterion allows to find the same interpretations as found with a visual scanning. A patient-specific system is able to set its optimal parameters according to the patient which makes it more accurate than non-patient-specific system. This paper introduces a new patient-specific system with genetic and practical swarm optimisation algorithms. The results show that the proposed system is able to reach acceptable performances. Moreover, the use of the genetic algorithm improves the system sensitivity (95%) more than the practical swarm optimization (91%) which makes it a better method for the system parameter optimisation.

Genetic and practical swarm optimisation algorithms for patient-specific seizure detection systems

The presence of artifacts in the EEG signals can cause a misunderstanding of the sought neurophysiological phenomena. In particular, the eye blink artifacts frequently contaminate the EEG and deteriorate its quality. Unfortunately, removing this artifact can lose useful information. The most popular approach in this field uses the independent component analysis to decompose the signal into different independent components and then eliminate those that are related to the artifact. Despite its reputation, this approach is computationally intensive, requires an acquisition from large number of channels and can alter the original EEG signal. In this paper, we propose a new method for blink artifact reduction. Several reference signals representing the eye blink are created then used in an orthogonal projection in order to cancel the artifact. The results of experiments, using 54 datasets from 27 subjects, show that the proposed method significantly outperformed the standard ADJUST, MARA and SASICA methods in removing the artifacts while preserving the pure EEG signals.

Omar Trigui

Papers

SSVEP Enhancement Using Moving Average Filter Controlled by Phase Features

Bispectral analysis-based approach for steady-state visual evoked potentials detection

A comparison study of SSVEP detection methods using the Emotiv Epoc headset

Genetic and practical swarm optimisation algorithms for patient-specific seizure detection systems

Removal of eye blink artifacts from EEG signal using morphological modeling and orthogonal projection