A. H. Jahidin

Classification of intelligence quotient via brainwave sub-band power ratio features and artificial neural network

This paper elaborates on the novel intelligence assessment method using the brainwave sub-band power ratio features. The study focuses only on the left hemisphere brainwave in its relaxed state. Distinct intelligence quotient groups have been established earlier from the score of the Raven Progressive Matrices. Sub-band power ratios are calculated from energy spectral density of theta, alpha and beta frequency bands. Synthetic data have been generated to increase dataset from 50 to 120. The features are used as input to the artificial neural network. Subsequently, the brain behaviour model has been developed using an artificial neural network that is trained with optimized learning rate, momentum constant and hidden nodes. Findings indicate that the distinct intelligence quotient groups can be classified from the brainwave sub-band power ratios with 100% training and 88.89% testing accuracies.

Brainwave sub-band power ratio characteristics in intelligence assessment

This paper discusses on the brainwave sub-band characteristics for different intelligence groups based on electroencephalogram (EEG) power ratio technique. The EEG datasets have been collected from 50 healthy subjects for two sessions; at relaxed, closed eye (CE) state as reference and at cognitively-stimulated state. In the stimulated state, subjects need to answer the intelligence quotient (IQ) test based on Ravens Standard Progressive Matrices (RPM). Sub-band power ratio from the two sessions were calculated and further analyzed to observe the pattern among different IQ groups. The results show that by implementing power ratio technique, the pattern of IQ groups, especially in the relaxed state can be clearly observed. It can be concluded that the value for alpha ratio is higher for high IQ group compared to low IQ group. In contrast to beta and theta ratio where high IQ groups have lower value compared to the low IQ group. This indicates that the ESD ratios can discriminate the characteristic of brainwaves for intelligence assessment.

Detection of cardiomyopathy using multilayered perceptron network

Cardiomyopathy refers to gradual weakening of the muscular walls of the cardiac chambers. Due to the hypertrophic condition of the muscular walls, damage and stretching of the muscle may lead to arrhythmias, which is detectable using the ECG. In the past, any deviations from a healthy rhythm provide cardiologists with accurate information regarding the heart condition. However, cardiologists are prone to making inaccurate interpretation from the visual observation, leading to erroneous diagnosis. Hence, this paper proposes a computerized method for accurate analysis and detection of cardiomyopathy disease using MLP network. Data for normal, cardiomyopathy, and other arrhythmias were obtained from the PTB Diagnostic ECG database. The raw signals were preprocessed for high-frequency noise removal using median and moving average filters. Baseline corrections were conducted using two-stage polynomial fitting method. Nine time-based features were extracted from the three bipolar limb leads. A total of 600 beats were used to train, validate and test five different MLP network structures. Four different learning algorithms were implemented to obtain the best classification accuracy and fastest convergence rate. Results show that the Levenberg-Marquardt algorithm shows the highest average classification accuracy of 98.9% for the different structures with the fastest average convergence rate of 12 epochs.

Identification of cardiomyopathy disease using hybrid multilayered perceptron network

Cardiomyopathy is a progressive disease that affects the muscular walls of the heart. The resultant hypertrophic condition of the cardiac chambers alters the capability of the heart to contract which will then lead to deterioration of cardiac output. The abnormality can manifest itself in the form of an arrhythmic signal detectable by electrocardiogram (ECG). Hence, this paper proposes the hybrid multilayered perceptron (HMLP) network for identification of cardiomyopathy disease. Initially, raw signals were acquired from the PTB Diagnostic ECG database for healthy, cardiomyopathy and other arrhythmias. The ECG underwent a signal preprocessing stage for noise reduction and baseline correction. Then, nine time-based sub-wave descriptors from the bipolar limb leads were retrieved via the median threshold approach. 600 beat samples were then utilized to train, test and validate the performance of the HMLP network. The HMLP network structures were tested for five variations of hidden nodes with four different learning algorithms. Findings indicate that the best convergence rate and detection accuracy are achievable with the Levenberg-Marquardt algorithm. Hence, the results suggest the potential application of HMLP for classification of arrhythmias.

Comparison between the Left and the Right Brainwaves for Delta and Theta Frequency Band after Horizontal Rotation Intervention

This paper investigates the difference between the left and the right brainwaves for delta and theta frequency band after horizontal rotation (HR) intervention using electroencephalography (EEG). The EEG signals were captured from the samples before and after under-going HR. The artifact of the EEG signals was removed automatically by means of a program designed to eliminate data above 100 micro volts. Then the Power Spectral Density (PSD) was applied to the artifact removed data and comparison was carried out to determine the better method and the effects of HR intervention. MATLAB and Paired T-test analysis from SPSS was used and graphs were plotted to show the correlation between the left and right brainwave before and after HR involvement. In conclusion, by applying PSD to the artifact removed signals, results improved significantly. It also shows that the correlation of the left and the right brainwaves had improved for delta and theta frequency bands after HR.

Robust arrhythmia classifier using hybrid multilayered perceptron network

The paper describes a robust approach to model cardiac arrhythmias using the hybrid multilayered perceptron (HMLP) network. Healthy, cardiomyopathy, as well as left and right bundle branch block electrocardiograms (ECG) was obtained from the PTB Diagnostic ECG database. The signals were initially pre-processed for noise removal and baseline correction. 24 morphological descriptors from the bipolar limb leads were used as input to the neural network. 400 beat samples were obtained for each condition. Results show that the Levenberg-Marquardt algorithm attains the fastest convergence. Varying the number of hidden nodes however, has no significant effect on the classification accuracy. Performance comparison shows that the HMLP network is more robust and gives better classification accuracy over the multilayered perceptron (MLP) network. The error convergence meanwhile, indicates a leveled performance.

Evaluation of brainwave sub-band spectral centroid in human intelligence

Sub-band spectral centroid (SC) has been widely applied in audio and speech processing field. This paper highlights the SC feature as a new approach to evaluate human intelligence quotient (IQ). The study focuses on resting EEG of the left brain hemisphere. The SC feature is derived from Discrete Fourier Transform (DFT) of electroencephalogram (EEG) signals. Sub-band SCs are obtained for delta, theta, alpha and beta frequency bands. IQ scores from the Raven Progressive Matrices (RPM) have been utilized to categorize dataset into three distinct groups. The SC features are then evaluated for significant pattern among the different intelligence levels. Results on theta and beta sub-bands indicate a trending pattern. Hence by implementing SC features of the theta and beta sub-bands, distinct IQ groups can be recognized.

Investigation on Elman neural network for detection of cardiomyopathy

Deterioration of structure and function of heart muscle is indicative of a degenerative disease known as cardiomyopathy. As a result, the hypertrophic condition of the heart often revealed itself in the form of abnormal sinus rhythm that can be detected via an electrocardiogram (ECG). In order to reduce the risk of misinterpretation by cardiologists, a variety of computational methods have been suggested for automated classification of arrhythmias. This paper proposes to explore Elman neural network for detecting cardiomyopathy. A total of 600 ECG beat samples were acquired from an established online database. Initially, the signals were filtered to eliminate high-frequency interference and perform baseline rectification. Nine time-based descriptors from leads I, II and III were used for training, testing and validation of the network structures. A total of five hidden-node node structures were tested with four different learning algorithms. Results show that all the network structure managed to achieve more than 90% classification accuracy. The fastest convergence was achieved with the Levenberg-Marquardt algorithm with an average of 16 epochs.

Principal component analysis and arrhythmia recognition using Elman neural network

Cardiac arrhythmia refers to any abnormal electrical activity in the heart that causes irregular heartbeat. Under clinical settings, the arrhythmias can be monitored non-invasively using the electrocardiogram (ECG). Although reliable, the method is still prone to error due to its dependence on visual interpretation. This paper proposes a computerized method for recognition of cardiac arrhythmia using Elman neural network. 1600 ECG beat samples for healthy, cardiomyopathy, and bundle branch block arrhythmias were acquired from the PTB Diagnostic ECG database. Initially, de-noising and baseline wander rectification were performed using digital filters and polynomial fitting technique. 24 morphological features from Lead I, II and III were obtained through the median threshold method. Principal component analysis was then implemented for feature selection. The dataset were reduced to 15 features and is then used to train, test and validate the Elman neural network structure with four different learning algorithms. The overall network performance is then benchmarked with the original 24 dataset. Results show that both settings attained classification accuracies of more than 95%. In addition, PCA has successfully reduced the feature requirements while simultaneously maintaining the network performance.

Classification of bundle branch blocks using multilayered perceptron network

Development of automated and accurate techniques for ECG recognition is important for diagnosis of heart diseases. Arrhythmic signals occur due to the disturbances to the rate, regularity, nodes and conduction path of the electrical impulses. Bundle branch block arises from defects of the conduction pathways involving blockage of electrical impulses through the bundle branches. This paper investigates MLP network for classification of bundle branch block arrhythmias. Trainings were conducted for varying network topologies with different training algorithms. A 98.2% overall detection accuracy was achieved over 90 beat samples. Results show that the Levenberg-Marquardt algorithm managed to achieve 100% recognition accuracy for all network topologies.