Mohd Afzan Othman

A new semantic mining approach for detecting ventricular tachycardia and ventricular fibrillation

Accurately differentiating between ventricular fibrillation (VF) and ventricular tachycardia (VT) episodes is crucial in preventing potentially fatal misinterpretations. If VT is misinterpreted as VF, the patient will receive an unnecessary shock that could damage the heart; conversely, if VF is incorrectly interpreted as VT, the result will be life-threatening. In this study, a new method called semantic mining is used to characterize VT and VF episodes by extracting their significant characteristics (the frequency, damping coefficient and input signal). This newly proposed method was tested using a widely recognized database provided by the Massachusetts Institute of Technology (MIT) and achieved high detection accuracy of 96.7%. The semantic mining technique was capable of completely discriminating between normal rhythms and VT and VF episodes without any false detections and also distinguished VT and VF episodes from one another with a recognition sensitivity of 94.1% and 95.2% for VT and VF, respectively.

Characterization of Ventricular Arrhythmias in Electrocardiogram Signal Using Semantic Mining Algorithm

Ventricular arrhythmias, especially ventricular fibrillation, is a type of arrhythmias that can cause sudden death. The paper applies semantic mining approach to electrocardiograph (ECG) signals in order to extract its significant characteristics (frequency, damping coefficient and input signal) to be used for classification purpose. Real data from an arrhythmia database are used after noise filtration. After features extraction they are statistically classified into three groups, i.e. normal (N), normal patients (PN) and patients with ventricular arrhythmia (V). We found that the V, PN, and N types of ECG signals can be identified by the extracted parameters. It is estimated that the parameters in semantic algorithm can be use to predict the onset of ventricular arrhythmias.

Dynamic ECG features for atrial fibrillation recognition

BACKGROUND Atrial fibrillation (AF) can cause the formation of blood clots in the heart. The clots may move to the brain and cause a stroke. Therefore, this study analyzed the ECG features of AF and normal sinus rhythm signals for AF recognition which were extracted by using a second-order dynamic system (SODS) concept. OBJECTIVE To find the appropriate windowing length for feature extraction based on SODS and to determine a machine learning method that could provide higher accuracy in recognizing AF. METHOD ECG features were extracted based on a dynamic system (DS) that uses a second-order differential equation to describe the short-term behavior of ECG signals according to the natural frequency (ω), damping coefficient, (ξ), and forcing input (u). The extracted features were windowed into 2, 3, 4, 6, 8, and 10 second episodes to find the appropriate windowing size for AF signal processing. ANOVA and t-tests were used to determine the significant features. In addition, pattern recognition machine learning methods (an artificial neural network (ANN) and a support vector machine (SVM)) with k-fold cross validation (k-CV) were used to develop the ECG recognition system. RESULTS Significant differences (p < 0.0001) were observed among all ECG groups (NSR, N, AF) using 2, 3, 4 and 6 second episodes for the features ω and u/ω; 4, 6 and 8 second episodes for features ω and u; 4 and 6 second episodes for features ω, u and u/ω, and; 10 second episodes for the feature ξ. The highest accuracy for AF recognition (AF, NSR) using ANN with k-CV was 95.3% using combination of features (ω and u; ω, u and u/ω) and SVM with k-CV was 95.0% using a combination of features ω, u and u/ω. CONCLUSION This study found that 4 s is the most appropriate windowing length, using two features (ω and u) for AF detection with an accuracy of 95.3%. Moreover, the pattern recognition learning machine uses an ANN with 10-fold cross validation based on DS.

CHARACTERIZATION OF VENTRICULAR ARRHYTHMIAS USING A SEMANTIC MINING ALGORITHM

Ventricular arrhythmia, especially ventricular fibrillation, is a type of arrhythmia that can cause sudden death. The aim of this paper is to characterize ventricular arrhythmias using semantic mining by extracting their significant characteristics (frequency, damping coefficient and input signal) from electrocardiogram (ECG) signals that represent the biological behavior of the cardiovascular system. Real data from an arrhythmia database are used after noise filtering and were statistically classified into two groups; normal sinus rhythm (N) and ventricular arrhythmia (V). The proposed method achieved high sensitivity and specificity (98.1% and 97.7%, respectively) and was capable of describing the differences between the N and V types in the ECG signal.

Characterization of Ventricular Tachycardia and Fibrillation Using Semantic Mining

Ventricular tachycardia (VT) and ventricular fibrillation (VF) are potentially life-threatening forms of cardiac arrhythmia. Fast and accurate detection of these conditions can save lives. We used semantic mining to characterize VT and VF episodes by extracting three significant parameters (frequency, damping coefficient and input signal) from electrocardiogram (ECG) signal. This method was used to analyze four-second ECG signals from a widely recognized database at the Massachusetts Institute of Technology (MIT). The method achieved a high sensitivity and specificity of 96.7% and 98.3%, respectively, and was capable of detecting normal sinus rhythm (N) from VT and VF signals without false detection, with a sensitivity of 100%. VT and VF signals were recognized from each other, with a recognition sensitivity of 96% and 94%, respectively. This newly proposed method using semantic mining shows strong potential for clinical applications because it is able to recognize VT and VF signals with higher accuracy and faster recognition times compare to existing methods.

Effect of ECG episodes on parameters extraction for paroxysmal atrial fibrillation classification

Atrial fibrillation is a type of atria arrhythmia which can cause the formation of blood clot in the heart. The blood clot may enlarge or moving to the brain and cause stroke. Therefore, this study monitors the performance of ECG episodes for paroxysmal atrial fibrillation classification. Episode of 2 seconds to 8 seconds were used to observe the performance of electrocardiograph (ECG) signal processing of atrial fibrillation patient classification. Methods of features extraction were based on the concept of describing short-term behaviour of complex physical and biological system, namely second order system (SOS), and with modified algorithm (hybrid with fast-Fourier transform, FFT). Features extracted from the ECG signal of atrial fibrillation patient were defined using three parameters, i.e. natural frequency, forcing input and damping coefficient. A total of twelve parameters were observed. Comparisons of performance between length of ECG episodes were explored for SOS, FFT-SOS and SOS-FFT algorithms. The episode of 4 seconds using SOS algorithm provides the highest accuracy (98 %) during the classification of ECG signal.

Atrial fibrillation classification and association between the natural frequency and the autonomic nervous system

BACKGROUND The feasibility study of the natural frequency (ω) obtained from a second-order dynamic system applied to an ECG signal was discovered recently. The heart rate for different ECG signals generates different ω values. The heart rate variability (HRV) and autonomic nervous system (ANS) have an association to represent cardiovascular variations for each individual. This study further analyzed the ω for different ECG signals with HRV for atrial fibrillation classification. METHODS This study used the MIT-BIH Normal Sinus Rhythm (nsrdb) and MIT-BIH Atrial Fibrillation (afdb) databases for healthy human (NSR) and atrial fibrillation patient (N and AF) ECG signals, respectively. The extraction of features was based on the dynamic system concept to determine the ω of the ECG signals. There were 35,031 samples used for classification. RESULTS There were significant differences between the N & NSR, N & AF, and NSR & AF groups as determined by the statistical t-test (p<0.0001). There was a linear separation at 0.4s(-1) for ω of both databases upon using the thresholding method. The feature ω for afdb and nsrdb falls within the high frequency (HF) and above the HF band, respectively. The feature classification between the nsrdb and afdb ECG signals was 96.53% accurate. CONCLUSIONS This study found that features of the ω of atrial fibrillation patients and healthy humans were associated with the frequency analysis of the ANS during parasympathetic activity. The feature ω is significant for different databases, and the classification between afdb and nsrdb was determined.

Artificial intelligence classification methods of atrial fibrillation with implementation technology

Abstract Background: Atrial fibrillation (AFIB) is one of the most common types of arrhythmia, which leads to heart failure and stroke to public. As AFIB has the high potential to cause permanent disability in patients, its early detection is extremely important. There are different types of AFIB classification algorithm that have been proposed by researchers in recent years. Methods: This paper reviews the features of AFIB in terms of ECG morphological features and heart rate variability (HRV) analysis on different methods. The existing classification method, particularly focusing on Artificial Intelligence technique, is also comprehensively described. Other than that, the existing implementation technology of arrhythmia detection platforms such as smart phone and System-on-Chip-based embedded device are also elaborated in terms of their design trade-offs. Conclusion: Current existing AFIB detection algorithm cannot compromise for high accuracy and low complexity. Due to the limitation of embedded system, design trade off should be considered to strike the balance between the performance of algorithm and the limitation.

ECG features extraction using second-order dynamic system and regeneration using hybrid recurrent network

ECG signals show the hearts condition for each individual. ECG signals characteristic can be extracted by using several methods such as P-wave conditions, RR-interval, fast-Fourier transform, wavelet transform, and etc. This study shows the relationship between features extraction of ECG signals by using second-order dynamic system (SODS) technique and ECG signals regeneration by using hybrid-recurrent network (HRN). HRN technique describes the mathematical proof of the algorithms used in SODS. The algorithm was developed by using Matlab software platform. Comparison was made and it was found that the ECG features extracted from SODS can be used to regenerate the ECG signals based on HRN technique. Therefore, the features extracted from SODS were valid to be used for further analysis of ECG signals.

Human Heart Oscillatory Behavior during Atrial Fibrillation Based on Second Order System

In this paper, the viability of a second order system to characterize the oscillatory behavior of human heart of atrial fibrillation patient was monitored and analysed. Sampling were patients who experienced atrial fibrillation. This study used the MIT-BIH Atrial Fibrillation dataset (MIT-afdb) from the Physiobank ECG database. ECG recordings of normal sinus rhythm (N) and atrial fibrillation (AF) which occurred sequentially, were analyzed for both ECG’s Lead I and Lead II. From here, the oscillatory behavior of human heart was characterized in accordance to the extracted parameters for each rhythm, i.e. natural frequency, damping coefficient and forcing input from the second order system. Results show that there were significant differences in mean value of natural frequency (ω), ratios of natural frequency to damping coefficient (ω/ζ), derivative of the natural frequency with respect to time (dω/dt) and derivative of the forcing input with respect to time (dμ/dt), between N and AF from Lead I (P < 0.01). Each parameter provides more than 95% accuracy using artificial neural networks.