Khaldon Lweesy

Automated sleep stage identification system based on time-frequency analysis of a single EEG channel and random forest classifier

In this work, an efficient automated new approach for sleep stage identification based on the new standard of the American academy of sleep medicine (AASM) is presented. The propose approach employs time-frequency analysis and entropy measures for feature extraction from a single electroencephalograph (EEG) channel. Three time-frequency techniques were deployed for the analysis of the EEG signal: Choi-Williams distribution (CWD), continuous wavelet transform (CWT), and Hilbert-Huang Transform (HHT). Polysomnographic recordings from sixteen subjects were used in this study and features were extracted from the time-frequency representation of the EEG signal using Renyis entropy. The classification of the extracted features was done using random forest classifier. The performance of the new approach was tested by evaluating the accuracy and the kappa coefficient for the three time-frequency distributions: CWD, CWT, and HHT. The CWT time-frequency distribution outperformed the other two distributions and showed excellent performance with an accuracy of 0.83 and a kappa coefficient of 0.76.

Classification of Sleep Stages Using Multi-wavelet Time Frequency Entropy and LDA

BACKGROUND The process of automatic sleep stage scoring consists of two major parts: feature extraction and classification. Features are normally extracted from the polysomnographic recordings, mainly electroencephalograph (EEG) signals. The EEG is considered a non-stationary signal which increases the complexity of the detection of different waves in it. OBJECTIVES This work presents a new technique for automatic sleep stage scoring based on employing continuous wavelet transform (CWT) and linear discriminant analysis (LDA) using different mother wavelets to detect different waves embedded in the EEG signal. METHODS The use of different mother wavelets increases the ability to detect waves in the EEG signal. The extracted features were formed based on CWT time frequency entropy using three mother wavelets, and the classification was performed using the linear discriminant analysis. Thirty-two data sets from the MIT-BIH database were used to evaluate the performance of the proposed method. RESULTS Features of a single EEG signal were extracted successfully based on the time frequency entropy using the continuous wavelet transform with three mother wavelets. The proposed method has shown to outperform the classification based on a CWT using a single mother wavelet. The accuracy was found to be 0.84, while the kappa coefficient was 0.78. CONCLUSIONS This work has shown that wavelet time frequency entropy provides a powerful tool for feature extraction for the non-stationary EEG signal; the accuracy of the classification procedure improved when using multiple wavelets compared to the use of single wavelet time frequency entropy.

A wireless home safety gas leakage detection system

A wireless safety device for gas leakage detection is proposed. The device is intended for use in household safety where appliances and heaters that use natural gas and liquid petroleum gas (LPG) may be a source of risk. The system also can be used for other applications in the industry or plants that depend on LPG and natural gas in their operations. The system design consists of two main modules: the detection and transmission module, and the receiving module. The detection and transmitting module detects the change of gas concentration using a special sensing circuit built for this purpose. This module checks if a change in concentration of gas(es) has exceeded a certain pre-determined threshold. If the sensor detects a change in gas concentration, it activates and audiovisual alarm and sends a signal to the receiver module. The receiver module acts as a mobile alarm device to allow the mobility within the house premises. The system was tested using LPG and the alarm was activated as a result of change in concentration.

Voiceless Arabic vowels recognition using facial EMG

This work attempts to recognize the Arabic vowels based on facial electromyograph (EMG) signals, to be used for people with speech impairment and for human computer interface. Vowels were selected since they are the most difficult letters to recognize by people in Arabic language. Twenty subjects (7 females and 13 males) were asked to pronounce three Arabic vowels continuously in a random order. Facial EMG signals were recorded over three channels from the three main facial muscles that are responsible for speech. The EMG signals are then pre-processed to eliminate noise and interference signals. Segmentation procedure was implemented to extract the time event that corresponds to each vowel based on a moving standard deviation window. The accuracy of the segmentation procedure was found to be 94%. The recognition of the vowels was carried out by extracting features from the EMG in three domains: the temporal, the spectral, and the time frequency using the wavelet packet transform. Classification of the extracted features was then finally performed using different classification methods implemented in the WEKA software. The random forest classifier with time frequency features showed the best performance with an accuracy of 77% evaluated using a 10-fold cross-validation.

Medical waste management practices in southern Jordan

The purpose of this work was to study medical waste practices in southern Jordan. Five major hospitals in that area were used to conduct this study. It involved distributing a modified version of the questionnaire that was adopted by the World Health Organization (WHO). The surveyed hospitals had total number of beds ranging from 72 to 131 beds, while the average amount of medical waste generated in these hospitals was 0.73 kg bed−1 day–1. The practices of medical waste management were found to be suffering from major defects owing to the lack of both training and proper education.

Noninvasive transdermal insulin delivery using ultrasound transducers

Noninvasive transdermal delivery of insulin is suggested in this paper using ultrasound transducers to improve the quality of life of diabetic patients. It is a preferable technique over the traditional invasive and painful subcutaneous insulin injections. Ten piston-shaped ultrasound transducers operating in the mid frequency range (100–200 kHz) were housed to include a reservoir that intended to hold insulin during in vivo transdermal delivery. Eleven rabbits were divided into two groups. The control group (n=6) did not receive ultrasound while the exposure group (n=5) received ultrasound for only ten minutes. Over the recording period of 60 minutes, blood Glucose levels in the control group remained around 125 mg/dl; while for the exposure group, it reduced from 132.4 mg/dl at the beginning of the experiment to 87.6 mg/dl after 60 minutes. Piston transducers in the mid range frequency were found feasible in transdermal insulin delivery in vivo using local rabbits. More investigations are required to test more frequency ranges with different hyperglycemic rabbit models.

New Automated Detection Method of OSA Based on Artificial Neural Networks Using P-Wave Shape and Time Changes

This paper describes a new method for automatic detection of obstructive sleep apnea (OSA) based on artificial neural networks (ANN) using regular electrocardiogram (ECG) recordings. ECG signals were pre-processed and segmented to extract the P-waves; then three P-wave features were extracted: the P-wave duration (Tp), the P-wave dispersion (Pd), and the time interval from the peak of the P-wave to the R-wave (Tpr). Combinations of the three features were used as features for classification using ANN. For each feature combination studied, 70% of the input data was used for training the ANN, 15% for validating, and 15% for testing the results. Perfect agreement between expert’s scores and the ANN scores was achieved when the ANN was applied on Tp, Pd, and Tpr taken together, while substantial agreements were achieved when applying the ANN on the feature combinations Tp and Pd, and Tp and Tpr.

Neonatal sleep state identification using deep learning autoencoders

Neonatal sleep state analysis provides a tool for diagnosis of several possible physiological disorders in newborns. The sleep state identification is a time consuming procedure where the sleep state is determined in epochs of 60 second for an entire sleep recording. A new technique for automated sleep state identification in neonates is proposed. The proposed system comprises two major step; feature extraction and classification. Twelve features were extracted from a single EEG recording. The features extracted were based on statistical parameters extracted from both temporal and spectral domains of the EEG signal. The total number of recordings used was 29 EEG recordings acquired from newborns (14 preterm infants and 15 fullterm). The classification was done based on deep autoencoder neural networks. The structure of the network used was two autoencoder layers and one softnet output layer. The performance of the proposed system was evaluated for both preterm and fullterm records assembled in one group of data using 10 fold cross validation. Also, the performance was tested for the two groups separately. The reported accuracy was 0.804 for the entire data sets.

Newborn sleep stage identification using multiscale entropy

Neonatal sleep stage identification is of great importance as it helps diagnosis of certain possible disabilities in newborns. The sleep stage identification is normally done manually for an entire sleep recording which requires great human resources; therefore a reliable automated sleep stage identification system offers a helpful tool for specialists. This study demonstrated a new method for automated sleep stage scoring in neonates. The automated approach comprises two major steps: feature extraction and classification. This study presented a new approach for feature extraction based on multiscale entropy (MSE), a recently developed method for the analysis of time series and physiological signals. The features were extracted from a single EEG recording where 13 recordings from preterm infants and 14 from full term infants were used. The classification was done using the Weka software with three different classifiers: neural networks, random forests, and classification via regression. The performance of the proposed method was found to be comparable to the methods reported in the literature. The reported accuracy was found to be 0.813 for preterm subjects and 0.864 for fullterm subjects.

Design, Construction, and Evaluation of an Electrical Impedance Myographer

This paper describes the design, construction, and evaluation of an electrical impedance myographer (EIM), which can be used as a non-invasive technique for the assessment of the muscular state. It can also be used for many diagnostic purposes such as distinguishing tumor tissue from normal tissue, estimating segmental muscle volume by bioelectrical impedance spectroscopy, predicting muscle mass and improving estimation of glomerular filtration rate in non-diabetic patients with chronic kidney disease. In the design described herein, two impedance spectra are generated: one for relaxed muscles and one for contracted muscles. Those two spectra are then compared to provide the evidence about the induced physiological modifications in muscle morphology. The EIM design consists of a Wien bridge oscillator that generates a 91 kHz sinusoidal signal, and a voltage to current converter which generates a constant current (1 mA) that passes through the patient’s forearm. An envelop detector was used to convert the signal from AC to DC. The output of the envelope was isolated from the next stage using a buffer. A low pass filter (0.7-20 Hz) was used to eliminate the high frequency noise and to get only low frequency signal. An optocoupler was used to ensure patient’s safety. From the output signal, it has been observed that the impedance of a healthy subject increases gradually as the muscle contracts; this is due to the increasing blood flow to the contracted muscle area. When the patient’s muscle was continuously contracted, the impedance signal appeared as DC line, and the signal decreased as the patient relaxed his muscle.