Syed Khairul Bashar

On the classification of sleep states by means of statistical and spectral features from single channel Electroencephalogram

Traditional sleep scoring based on visual inspection of Electroencephalogram (EEG) signals is onerous for sleep scorers because of the gargantuan volume of data that have to be analyzed per examination. Computer-aided sleep staging can alleviate the onus of the sleep scorers. Again, most of the existing works on automatic sleep staging are multichannel based. Multichannel based sleep scoring is not pragmatic for the implementation of a wearable and portable sleep quality evaluation device. Due to all these factors, automatic sleep scoring based on single channel EEG is garnering increasing attention of sleep researchers. In this work, we propound a single channel based solution to sleep scoring. First, we decompose the EEG signals into segments. We then compute various statistical and spectral features from the signal segments. After performing statistical analyses, we perform classification using artificial neural network. Results of various experiments perspicuously manifest that the proposed scheme is superior to state-of-the-art ones in accuracy.

Automatic classification of sleep stages from single-channel electroencephalogram

A portable and wearable yet low-power sleep monitoring system necessitates an automatic sleep scoring algorithm with the use of minimum number of recording channels. Computer-aided sleep staging is also important to eradicate the onus of sleep scorers of analyzing an enormous volume of data. The existing works on sleep scoring are either multichannel based or yield poor performance. Therefore, an automatic sleep scoring algorithm based on single channel EEG signals is yet to emerge. In this work, we utilize spectral features to extract discriminatory information from EEG signal segments. We then perform statistical analyses to find out the efficacy and the discriminatory capability of the selected features for various sleep states. Afterwards, we employ Adaptive Boosting to perform classification. The experimental outcomes perspicuously manifest that the proposed scheme is superior to state-of-the-art ones in accuracy.

Identification of motor imagery movements from EEG signals using Dual Tree Complex Wavelet Transform

In this paper, Dual Tree Complex Wavelet Transform (DTCWT) domain based feature extraction method has been proposed to identify left and right hand motor imagery movements from electroencephalogram (EEG) signals. After first performing auto-correlation of the EEG signals to enhance the weak brain signals and reduce noise, the EEG signals are decomposed into several bands of real and imaginary coefficients using DTCWT. The energy of the coefficients from relevant bands have been extracted as features and from the one way ANOVA analysis, scatter plots, box plots and histograms, this features are shown to be promising to distinguish various kinds of EEG signals. Publicly available benchmark BCI-competition 2003 Graz motor imagery dataset is used for this experiment. Among different types of classifiers developed such as support vector machine (SVM), probabilistic neural network (PNN), adaptive neuro fuzzy inference system (ANFIS) and K-nearest neighbor (KNN), KNN classifiers have been shown to provide a good mean accuracy of 91.07% which is better than several existing techniques.

A statistical feature based novel method to detect bleeding in wireless capsule endoscopy images

Wireless capsule endoscopy (WCE) is a recently developed technology to detect small intestine diseases, such as bleeding. In this paper, a scheme for automatic bleeding detection from WCE video is proposed based on different statistical measures computed from a new red to green (R/G) pixel ratio intensity plane of RGB color images. Different statistical parameters, namely mean, mode, maximum, minimum, skewness, median, variance, and kurtosis are used to extract variation in spatial characteristics in R/G intensity plane of bleeding and non-bleeding WCE RGB images. Depending on the ability to provide significantly distinguishable characteristics, in the proposed feature vector, median, variance, and kurtosis of R/G ratio values corresponding to a WCE image are considered. For the purpose of classification, K-nearest neighbor (KNN) classifier is employed. From extensive experimentation on several WCE videos collected from a publicly available database, it is observed that the proposed method can successfully detect bleeding and non-bleeding images with high level of accuracy, sensitivity and specificity in comparison to that of some of the existing methods.

A feature extraction scheme from region of interest of wireless capsule endoscopy images for automatic bleeding detection

Wireless capsule endoscopy (WCE) is an effective but painless video technology to detect small intestine diseases like bleeding. For analyzing WCE video frames, instead of using the most common RGB (red, green, blue) color scheme, in this paper, CMYK (Cyan-C, Magenta-M, Yellow-I and Black-K) color scheme is used, which is subtractive color model and more effective for color separation. First, a region of interest (ROI) is determined using YIQ (luminance-Y, chrominance-IQ: in phase-I and quadrature-Q) color scheme depending on the Q value of the pixels and some morphological operations. Next, CMYK values are calculated within the ROI pixels. Instead of considering single color space all color spaces are investigated to extract feature, among them four statistical measures as mean of four color space is proposed. It is shown that use of ROI and CMYK color space not only reduces computational complexity but also offers significantly better discrimination between bleeding and non-bleeding pixels. For the purpose of classification, support vector machine (SVM) classifier is employed. From extensive experimentation on several WCE videos collected from a publicly available database, it is observed that the bleeding detection performance of the proposed method in terms of accuracy, sensitivity and specificity is quite satisfactory in comparison to that obtained by some of the existing methods.

Motor imagery movements classification using multivariate EMD and short time Fourier transform

In this paper, a novel method consisting of multivariate empirical mode decomposition (MEMD) and short time Fourier transform (STFT) is proposed to identify left and right imagery hand movements from electroencephalogram (EEG) signals. Experiments are carried out using the publicly available benchmark BCI competition 2003 Graz motor imagery data base of left and right hand movements. The EEG epochs are decomposed into multiple intrinsic mode functions (IMFs) by applying MEMD. The most significant mode is subjected to the short time Fourier transform; the peak and entropy of the magnitude spectrum are used as features representing the corresponding epoch. Extensive analysis is carried out using Kruskal-Wallis test, scatter plots, box plots and histograms to justify the employed features. Classification of the motor imagery movements is studied using the proposed features and various state of the art classifiers. The highest accuracy is achieved employing the k-nearest neighbor (kNN) classifier which is 90.00% and better than those of the several contemporary methods.

An automatic bleeding detection technique in wireless capsule endoscopy from region of interest

Wireless capsule endoscopy (WCE) is a painless operative video technology to detect small intestine diseases, such as bleeding. Instead of using the most common RGB (red, green, blue) color scheme, in this paper, YIQ (luminance-Y, chrominance-IQ: in phase-I and quadrature-Q) color scheme is used for analyzing WCE video frames, which corresponds better to human color response characteristics. Analyzing the behavior of each of the four YIQ spaces, first, a region of interest is determined depending on the Q value of the pixels and some morphological operations. Next, instead of considering three spaces of YIQ color model separately, a new composite space Y.I/Q is proposed to capture intrinsic information about the luminance and chrominance of images. Four statistical measures, namely mean, median, skewness and minima of the pixel values in composite space within the ROI are computed as features. It is exhibited that use of composite space lower computational complexity as well as offers noticeably better discrimination between bleeding and non-bleeding pixels. For the purpose of classification, support vector machine (SVM) classifier is employed. Satisfactory bleeding detection performance result is achieved in terms of accuracy, sensitivity and specificity from severe experimentation on several WCE videos which is collected from a publicly available database. Also it is observed that proposed method over performs with comparing some of the existing methods.

Dual tree complex wavelet transform domain based motor imagery movements classification scheme from EEG signals

In this paper, a feature extraction method based on Dual Tree Complex Wavelet Transform (DTCWT) domain has been proposed to classify left and right hand motor imagery movements from electroencephalogram (EEG) signals. After first performing auto-correlation of the EEG signals to reduce noise and enhance the weak brain signals, the EEG signals are decomposed into several bands of real and imaginary coefficients using DTCWT. Then mean absolute deviation has been extracted as features from relevant coefficient bands and from the one way ANOVA analysis, scatter plots, box plots and histograms, this feature has shown to be promising to distinguish various kinds of EEG signals. The total experiment is carried out with the publicly available benchmark BCI-competition 2003 Graz motor imagery data set. Among different types of classifiers used K-nearest neighbor (KNN)-based classifiers have been shown to provide a good mean accuracy of 90.36% which is better than several existing techniques.

Motor imagery movements detection of EEG signals using statistical features in the Dual Tree Complex Wavelet Transform domain

In this paper, a statistical method has been proposed to identify motor imagery left and right hand movements from electroencephalogram (EEG) signals in the Dual Tree Complex Wavelet Transform (DTCWT) domain. The total experiment is carried out with the publicly available benchmark BCI-competition 2003 Graz motor imagery dataset. First, the EEG signals are decomposed into several bands of real and imaginary coefficients, and then, some statistical features like norm entropy and standard deviation have been calculated. From the one way ANOVA analysis, these features have been shown to be promising to distinguish various kinds of EEG signals. Various types of classifiers have been developed to realize the discrimination among the EEG signals. Among various types of classifiers, K-nearest neighbor (KNN)-based classifiers have been shown to provide a good accuracy of 90.36% which is shown to be better than several existing techniques.

Identification of arm movements using statistical features from EEG signals in wavelet packet domain

In this paper, a method to classify arm movements using statistical features of electroencephalogram (EEG) signals calculated from wavelet packet and Fourier transforms, has been proposed. The EEG signals are analyzed using bi-orthogonal wavelet packet family. Fourier transform is then applied to the corresponding detail coefficients and higher order statistical moment named kurtosis is calculated from the magnitude of the Fourier components. The features are shown to be distinguishable for the EEG signals of four different arm movements. K-nearest neighbor (KNN)-based classifiers are developed using these features to identify the arm movements, right hand forward and backward; left hand forward and backward. A mean accuracy of 92.84% is achieved which is shown to be better than some existing techniques.