Supervised dictionary learning of EEG signals for mild cognitive impairment diagnosis

Selecting EEG channels and features using multi-objective optimization for accurate MCI detection: validation using leave-one-subject-out strategy

Effective management of dementia requires the timely detection of mild cognitive impairment (MCI). This paper introduces a multi-objective optimization approach for selecting EEG channels (and features) for the purpose of detecting MCI. Firstly, each EEG signal from each channel is decomposed into subbands using either variational mode decomposition (VMD) or discrete wavelet transform (DWT). A feature is then extracted from each subband using one of the following measures: standard deviation, interquartile range, band power, Teager energy, Katz's and Higuchi's fractal dimensions, Shannon entropy, sure entropy, or threshold entropy. Different machine learning techniques are used to classify the features of MCI cases from those of healthy controls. The classifier's performance is validated using leave-one-subject-out (LOSO) cross-validation (CV). The non-dominated sorting genetic algorithm (NSGA)-II is designed with the aim of minimizing the number of EEG channels (or features) and maximizing classification accuracy. The performance is evaluated using a publicly available online dataset containing EEGs from 19 channels recorded from 24 participants. The results demonstrate a significant improvement in performance when utilizing the NSGA-II algorithm. By selecting only a few appropriate EEG channels, the LOSO CV-based results show a significant improvement compared to using all 19 channels. Additionally, the outcomes indicate that accuracy can be further improved by selecting suitable features from different channels. For instance, by combining VMD and Teager energy, the SVM accuracy obtained using all channels is 74.24%. Interestingly, when only five channels are selected using NSGA-II, the accuracy increases to 91.56%. The accuracy is further improved to 95.28% when using only 8 features selected from 7 channels. This demonstrates that by choosing informative features or channels while excluding noisy or irrelevant information, the impact of noise is reduced, resulting in improved accuracy. These promising findings indicate that, with a limited number of channels and features, accurate diagnosis of MCI is achievable, which opens the door for its application in clinical practice.

Spectral analysis and Bi-LSTM deep network-based approach in detection of mild cognitive impairment from electroencephalography signals

Mild cognitive impairment (MCI) is a neuropsychological syndrome that is characterized by cognitive impairments. It typically affects adults 60 years of age and older. It is a noticeable decline in the cognitive function of the patient, and if left untreated it gets converted to Alzheimer's disease (AD). For that reason, early diagnosis of MCI is important as it slows down the conversion of the disease to AD. Early and accurate diagnosis of MCI requires recognition of the clinical characteristics of the disease, extensive testing, and long-term observations. These observations and tests can be subjective, expensive, incomplete, or inaccurate. Electroencephalography (EEG) is a powerful choice for the diagnosis of diseases with its advantages such as being non-invasive, based on findings, less costly, and getting results in a short time. In this study, a new EEG-based model is developed which can effectively detect MCI patients with higher accuracy. For this purpose, a dataset consisting of EEG signals recorded from a total of 34 subjects, 18 of whom were MCI and 16 control groups was used, and their ages ranged from 40 to 77. To conduct the experiment, the EEG signals were denoised using Multiscale Principal Component Analysis (MSPCA), and to increase the size of the dataset Data Augmentation (DA) method was performed. The tenfold cross-validation method was used to validate the model, moreover, the power spectral density (PSD) of the EEG signals was extracted from the EEG signals using three spectral analysis methods, the periodogram, welch, and multitaper. The PSD graphs of the EEG signals showed signal differences between the subjects of control and the MCI group, indicating that the signal power of MCI patients is lower compared to control groups. To classify the subjects, one of the best classifiers of deep learning algorithms called the Bi-directional long-short-term-memory (Bi-LSTM) was used, and several machine learning algorithms, such as decision tree (DT), support vector machine (SVM), and k-nearest neighbor (KNN). These algorithms were trained and tested using the extracted feature vectors from the control and the MCI groups. Additionally, the values of the coefficient matrix of those algorithms were compared and evaluated with the performance evaluation matrix to determine which one performed the best overall. According to the experimental results, the proposed deep learning model of multitaper spectral analysis approach with Bi-LSTM deep learning algorithm attained the highest number of correctly classified samples for diagnosing MCI patients and achieved a remarkable accuracy compared to the other proposed models. The achieved classification results of the deep learning model are reported to be 98.97% accuracy, 98.34% sensitivity, 99.67% specificity, 99.70% precision, 99.02% f1 score, and 97.94% Matthews correlation coefficient (MCC).

A tensor decomposition scheme for EEG-based diagnosis of mild cognitive impairment

Mild Cognitive Impairment (MCI) is the primary stage of acute Alzheimer's disease, and early detection is crucial for the person and those around him. It is difficult to recognize since this mild stage does not have clear clinical signs, and its symptoms are between normal aging and severe dementia. Here, we propose a tensor decomposition-based scheme for automatically diagnosing MCI using Electroencephalogram (EEG) signals. A new projection is proposed, which preserves the spatial information of the electrodes to construct a data tensor. Then, using parallel factor analysis (PARAFAC) tensor decomposition, the features are extracted, and a support vector machine (SVM) is used to discriminate MCI from normal subjects. The proposed scheme was tested on two different datasets. The results showed that the tensor-based method outperformed conventional methods in diagnosing MCI with an average classification accuracy of 93.96% and 78.65% for the first and second datasets, respectively. Therefore, it seems that maintaining the spatial topology of the signals plays a vital role in the processing of EEG signals.

Application of machine learning in dementia diagnosis: A systematic literature review

According to the World Health Organization forecast, over 55 million people worldwide have dementia, and about 10 million new cases are detected yearly. Early diagnosis is essential for patients to plan for the future and deal with the disease. Machine Learning algorithms allow us to solve the problems associated with early disease detection. This work attempts to identify the current relevance of the application of machine learning in dementia prediction in the scientific world and suggests open fields for future research. The literature review was conducted by combining bibliometric and content analysis of articles originating in a period of 20 years in the Scopus database. Twenty-seven thousand five hundred twenty papers were identified firstly, of which a limited number focused on machine learning in dementia diagnosis. After the exclusion process, 202 were selected, and 25 were chosen for analysis. The recent increasing interest in the past five years in the theme of machine learning in dementia shows that it is a relevant field for research with still open questions. The methods used to identify dementia or what features are used to identify or predict this disease are explored in this study. The literature review revealed that most studies used magnetic resonance imaging (MRI) and its types as the main feature, accompanied by demographic data such as age, gender, and the mini-mental state examination score (MMSE). Data are usually acquired from the Alzheimer's Disease Neuroimaging Initiative (ADNI). Classification of Alzheimer's disease is more prevalent than prediction of Mild Cognitive Impairment (MCI) or their combination. The authors preferred machine learning algorithms such as SVM, Ensemble methods, and CNN because of their excellent performance and results in previous studies. However, most use not one machine-learning technique but a combination of techniques. Despite achieving good results in the studies considered, there are new concepts for future investigation declared by the authors and suggestions for improvements by employing promising methods with potentially significant results.

PLI-Based Connectivity in Resting-EEG is a Robust and Generalizable Feature for Detecting MCI and AD: A Validation on a Diverse Multisite Clinical Dataset

The high prevalence rate of Alzheimer's disease (AD) and mild cognitive impairment (MCI) has been a serious public health threat to the modern society. Recently, many studies have demonstrated the potential of using non-invasive electroencephalography (EEG) and machine learning to assist the diagnosis of AD/MCI. However, the majority of these research recorded EEG signals from a single center, leading to significant concerns regarding the generalizability of the findings in clinical settings. The current study aims to reevaluate the effectiveness of EEG-based machine learning model for the detection of AD/MCI in the case of a relatively large and diverse data set. We collected resting-state EEG data from 150 participants across six hospitals and examined the classification performances of Linear Discriminative Analysis (LDA) classifiers on the phase lag index (PLI) feature. We also compared the performance of PLI over the other commonly-used EEG features and other classifiers. The model was first tested on a training set to select the feature subset and then further validated with an independent test set. The results demonstrate that PLI performs the best compared to other features. The LDA classifier trained with the optimal PLI features can provide 82.50% leave-one-participant-out cross-validation (LOPO-CV) accuracy on the training set and maintain a good enough performance with 75.00% accuracy on the test set. Our results suggest that PLI-based functional connectivity could be considered as a reliable bio-maker to detect AD/MCI in the real-world clinical settings.

Fault detection and analysis for wheelset bearings via improved explicit shift-invariant dictionary learning

The wheelset bearing is an indispensable part of the high-speed train, and monitoring its service performance is a concern of many researchers. Effective extraction of those impulse signals induced by the defects on the bearing elements is the key to fault detection and behaviour analysis. However, the presence of considerable noise and irrelevant components brings difficulties to extracting the wheelset bearing fault impulse signals from the measured vibration signals. This paper proposes an improved explicit shift-invariant dictionary learning (IE-SIDL) method to address this issue. Based on the shift-invariant characteristics of the wheelset bearing fault impulse signal in the time-domain, the circulant matrix is used to construct a shift-invariant dictionary and explicitly characterize the fault impulses at any time. To improve the efficiency of dictionary learning, a method of three flips is introduced to realize fast dictionary construction, and the frequency-domain reconstruction property of the circulant matrix is employed to quickly update the dictionary. Besides, an indicator-guided subspace pursuit (SP) method based on the sparsity of envelope spectrum (SES) is adopted for the sparse coding to improve sparse solution accuracy and adaptation. The effectiveness of the IE-SIDL method is proved through the simulated and experimental signals. The results demonstrate that the improved dictionary learning method has an excellent capacity in extracting fault impulse signal of the wheelset bearings, and the good time- and frequency-domain characteristics of the processed signals facilitate fault detection and behaviour analysis.

Which Multivariate Multi-Scale Entropy Algorithm Is More Suitable for Analyzing the EEG Characteristics of Mild Cognitive Impairment?

So far, most articles using the multivariate multi-scale entropy algorithm mainly use algorithms to analyze the multivariable signal complexity without clearly describing what characteristics of signals these algorithms measure and what factors affect these algorithms. This paper analyzes six commonly used multivariate multi-scale entropy algorithms from a new perspective. It clarifies for the first time what characteristics of signals these algorithms measure and which factors affect them. It also studies which algorithm is more suitable for analyzing mild cognitive impairment (MCI) electroencephalograph (EEG) signals. The simulation results show that the multivariate multi-scale sample entropy (mvMSE), multivariate multi-scale fuzzy entropy (mvMFE), and refined composite multivariate multi-scale fuzzy entropy (RCmvMFE) algorithms can measure intra- and inter-channel correlation and multivariable signal complexity. In the joint analysis of coupling and complexity, they all decrease with the decrease in signal complexity and coupling strength, highlighting their advantages in processing related multi-channel signals, which is a discovery in the simulation. Among them, the RCmvMFE algorithm can better distinguish different complexity signals and correlations between channels. It also performs well in anti-noise and length analysis of multi-channel data simultaneously. Therefore, we use the RCmvMFE algorithm to analyze EEG signals from twenty subjects (eight control subjects and twelve MCI subjects). The results show that the MCI group had lower entropy than the control group on the short scale and the opposite on the long scale. Moreover, frontal entropy correlates significantly positively with the Montreal Cognitive Assessment score and Auditory Verbal Learning Test delayed recall score on the short scale.

Automatic Diagnosis of Mild Cognitive Impairment Based on Spectral, Functional Connectivity, and Nonlinear EEG-Based Features

Accurate and early diagnosis of mild cognitive impairment (MCI) is necessary to prevent the progress of Alzheimer's and other kinds of dementia. Unfortunately, the symptoms of MCI are complicated and may often be misinterpreted as those associated with the normal ageing process. To address this issue, many studies have proposed application of machine learning techniques for early MCI diagnosis based on electroencephalography (EEG). In this study, a machine learning framework for MCI diagnosis is proposed in this study, which extracts spectral, functional connectivity, and nonlinear features from EEG signals. The sequential backward feature selection (SBFS) algorithm is used to select the best subset of features. Several classification models and different combinations of feature sets are measured to identify the best ones for the proposed framework. A dataset of 16 and 18 EEG data of normal and MCI subjects is used to validate the proposed system. Metrics including accuracy (AC), sensitivity (SE), specificity (SP), F1-score (F1), and false discovery rate (FDR) are evaluated using 10-fold crossvalidation. An average AC of 99.4%, SE of 98.8%, SP of 100%, F1 of 99.4%, and FDR of 0% have been provided by the best performance of the proposed framework using the linear support vector machine (LSVM) classifier and the combination of all feature sets. The acquired results confirm that the proposed framework provides an accurate and robust performance for recognizing MCI cases and outperforms previous approaches. Based on the obtained results, it is possible to be developed in order to use as a computer-aided diagnosis (CAD) tool for clinical purposes.

A Multichannel Intraluminal Impedance Gastroesophageal Reflux Characterization Algorithm Based On Sparse Representation

Gastroesophageal reflux disease (GERD) is a common digestive disorder with troublesome symptoms that has been affected millions of people worldwide. Multichannel Intraluminal Impedance-pH (MII-pH) monitoring is a recently developed technique, which is currently considered as the gold standard for the diagnosis of GERD. In this paper, we address the problem of characterizing gastroesophageal reflux events in MII signals. A GER detection algorithm has been developed based on the sparse representation of local segments. Two dictionaries are trained using the online dictionary learning approach from the distal impedance data of selected patches of GER and no specific patterns intervals. A classifier is then designed based on the l_p-norm of dictionary approximations. Next, a preliminary permutation mask is obtained from the classification results of patches, which is then used in post-processing procedure to investigate the exact timings of GERs at all impedance sites. Our algorithm was tested on 33 MII episodes, resulting a sensitivity of 96.97% and a positive predictive value of 94.12%.

A Hierarchical Discriminative Sparse Representation Classifier for EEG Signal Detection

Classification of electroencephalogram (EEG) signal data plays a vital role in epilepsy detection. Recently sparse representation-based classification (SRC) methods have achieved the good performance in EEG signal automatic detection, by which the EEG signals are sparsely represented using a few active coefficients in the dictionary and classified according to the reconstruction criteria. However, most of SRC learn a linear dictionary for encoding, and cannot extract enough information and nonlinear relationship of data for classification. To solve this problem, a hierarchical discriminative sparse representation classification model (called HD-SRC) for EEG signal detection is proposed. Based on the framework of neural network, HD-SRC learns the hierarchical nonlinear transformation and maps the signal data into the nonlinear transformed space. Through incorporating this idea into label consistent K singular value decomposition (LC-KSVD) at the top layer of neural network, HD-SRC seeks discriminative representation together with dictionary, while minimizing errors of classification, reconstruction and discriminative sparse-code for pattern classification. By learning the hierarchical feature mapping and discriminative dictionary simultaneously, more discriminative information of data can be exploited. In the experiment the proposed model is evaluated on the Bonn EEG database, and the results show it obtains satisfactory classification performance in multiple EEG signal detection tasks.

Identifying Individuals With Mild Cognitive Impairment Using Working Memory-Induced Intra-Subject Variability of Resting-State EEGs

Individuals with mild cognitive impairment (MCI) are at high risk of developing into dementia (e. g., Alzheimer's disease, AD). A reliable and effective approach for early detection of MCI has become a critical challenge. Although compared with other costly or risky lab tests, electroencephalogram (EEG) seems to be an ideal alternative measure for early detection of MCI, searching for valid EEG features for classification between healthy controls (HCs) and individuals with MCI remains to be largely unexplored. Here, we design a novel feature extraction framework and propose that the spectral-power-based task-induced intra-subject variability extracted by this framework can be an encouraging candidate EEG feature for the early detection of MCI. In this framework, we extracted the task-induced intra-subject spectral power variability of resting-state EEGs (as measured by a between-run similarity) before and after participants performing cognitively exhausted working memory tasks as the candidate feature. The results from 74 participants (23 individuals with AD, 24 individuals with MCI, 27 HC) showed that the between-run similarity over the frontal and central scalp regions in the HC group is higher than that in the AD or MCI group. Furthermore, using a feature selection scheme and a support vector machine (SVM) classifier, the between-run similarity showed encouraging leave-one-participant-out cross-validation (LOPO-CV) classification performance for the classification between the MCI and HC (80.39%) groups and between the AD vs. HC groups (78%), and its classification performance is superior to other widely-used features such as spectral powers, coherence, and the complexity estimated by Katz's method extracted from single-run resting-state EEGs (a common approach in previous studies). The results based on LOPO-CV, therefore, suggest that the spectral-power-based task-induced intra-subject EEG variability extracted by the proposed feature extraction framework has the potential to serve as a neurophysiological feature for the early detection of MCI in individuals.

Evaluation of the Brain Function State During Mild Cognitive Impairment Based on Weighted Multiple Multiscale Entropy

The mild cognitive impairment (MCI) stage plays an essential role in preventing the progression of older adults to Alzheimer's disease. In this study, neurofeedback training (NFT) is applied to improve MCI brain cognitive function. To assess the improvement effect, a novel algorithm called Weighted Multiple Multiscale Entropy (WMMSE) is proposed to extract and analyze the electroencephalogram (EEG) features of patients with MCI. To overcome the information loss problem of traditional multiscale entropy (MSE), WMMSE fully considered the correlation of the sequence and the contribution of each sequence to the total entropy. The experimental group composed of 39 patients with MCI was subjected to NFT for 10 days during two sessions. The control group included 21 patients with MCI without any intervention. The Lempel-Ziv complexity (LZC) was used for primary assessment, and WMMSE was used to accurately analyze the effect of NFT. The results show that the WMMSE values of F4, C3, C4, O1, and T5 channels post-NFT are higher compared with pre-NFT and significant differences (P < 0.05). Moreover, the cognitive subscale of the Montreal Cognitive Assessment (MoCA) results shows that the post-NFT score is higher than the pre-NFT in the vast majority of the patients with MCI and significant differences (P < 0.05). When compared with the control group, the WMMSE values of the experimental group increased in each channel. Therefore, the NFT intervention method contributes to brain cognitive functional recovery, and WMMSE can be used as a biomarker to evaluate the state of MCI brain cognitive function.

MCI Detection Using Kernel Eigen-Relative-Power Features of EEG Signals

Classification between individuals with mild cognitive impairment (MCI) and healthy controls (HC) based on electroencephalography (EEG) has been considered a challenging task to be addressed for the purpose of its early detection. In this study, we proposed a novel EEG feature, the kernel eigen-relative-power (KERP) feature, for achieving high classification accuracy of MCI versus HC. First, we introduced the relative powers (RPs) between pairs of electrodes across 21 different subbands of 2-Hz width as the features, which have not yet been used in previous MCI-HC classification studies. Next, the Fisher’s class separability criterion was applied to determine the best electrode pairs (five electrodes) as well as the frequency subbands for extracting the most sensitive RP features. The kernel principal component analysis (kernel PCA) algorithm was further performed to extract a few more discriminating nonlinear principal components from the optimal RPs, and these components form a KERP feature vector. Results carried out on 51 participants (24 MCI and 27 HC) show that the newly introduced subband RP feature showed superior classification performance to commonly used spectral power features, including the band power, single-electrode relative power, and also the RP based on the conventional frequency bands. A high leave-one-participant-out cross-validation (LOPO-CV) classification accuracy 86.27% was achieved by the RP feature, using a simple linear discriminant analysis (LDA) classifier. Moreover, with the same classifier, the proposed KERP further improved the accuracy to 88.24%. Finally, cascading the KERP feature to a nonlinear classifier, the support vector machine (SVM), yields a high MCI-HC classification accuracy of 90.20% (sensitivity = 87.50% and specificity = 92.59%). The proposed method demonstrated a high accuracy and a high usability (only five electrodes are required), and therefore, has great potential to further develop an EEG-based computer-aided diagnosis system that can be applied for the early detection of MCI.

Optimized Projection and Fisher Discriminative Dictionary Learning for EEG Emotion Recognition

Electroencephalogram (EEG)-based emotion recognition (ER) has drawn increasing attention in the brain–computer interface (BCI) due to its great potentials in human–machine interaction applications. According to the characteristics of rhythms, EEG signals usually can be divided into several different frequency bands. Most existing methods concatenate multiple frequency band features together and treat them as a single feature vector. However, it is often difficult to utilize band-specific information in this way. In this study, an optimized projection and Fisher discriminative dictionary learning (OPFDDL) model is proposed to efficiently exploit the specific discriminative information of each frequency band. Using subspace projection technology, EEG signals of all frequency bands are projected into a subspace. The shared dictionary is learned in the projection subspace such that the specific discriminative information of each frequency band can be utilized efficiently, and simultaneously, the shared discriminative information among multiple bands can be preserved. In particular, the Fisher discrimination criterion is imposed on the atoms to minimize within-class sparse reconstruction error and maximize between-class sparse reconstruction error. Then, an alternating optimization algorithm is developed to obtain the optimal solution for the projection matrix and the dictionary. Experimental results on two EEG-based ER datasets show that this model can achieve remarkable results and demonstrate its effectiveness.

Mixture Kernel Density Estimation and Remedied Correlation Matrix on the EEG-Based Copula Model for the Assessment of Visual Discomfort

Since electroencephalogram (EEG) signals can directly provide information on changes in brain activity due to behaviour changes, how to assess visual discomfort through EEG signals attracts researchers’ attention. However, previous assessments based on time-domain EEG features lack sufficient consideration of the dependence among EEG signals, which may affect the discrimination to visual discomfort. Although the copula model can explore the dependence among variables, the EEG-based copula models still have the following deficiencies: (1) the methods ignoring the fine-grained information hidden in EEG signals could make the estimated marginal density function improper, and (2) the approaches neglecting the pseudo-correlation among data may inappropriately estimate the correlation matrix parameter of the copula density function. The mixture kernel density estimation (MKDE) and remedied correlation matrix (RCM) on the EEG-based copula model are proposed to mitigate the mentioned shortcomings. The simulation experiments show that MKDE can not only better estimate the marginal density function but also explore fine-grained information. The RCM can be closer to the real correlation matrix parameter. With the favourable quality of the proposed EEG-based model, it is used to extract time-domain EEG features to assess visual discomfort further. To our best knowledge, the extracted features present better discrimination to visual discomfort compared with the features extracted by the state-of-the-art method.