The time-sequenced adaptive filter for analysis of cardiac arrhythmias in intraventricular electrograms

Reconstructed State Space Features for Classification of ECG Signals

Background:

Cardiac arrhythmias are considered as one of the most serious health conditions; therefore, accurate and quick diagnosis of these conditions is highly paramount for the electrocardiogram (ECG) signals. Moreover, are rather difficult for the cardiologists to diagnose with unaided eyes due to a close similarity of these signals in the time domain.

Objective:

In this paper, an image-based and machine learning method were presented in order to investigate the differences between the three cardiac arrhythmias of VF, VT, SVT and the normal signal.

Material and Methods:

In this simulation study, the ECG data used are collected from 3 databases, including Boston Beth University Arrhythmias Center, Creighton University, and MIT-BIH. The proposed algorithm was implemented using MATLAB R2015a software and its simulation. At first, the signal is transmitted to the state space using an optimal time delay. Then, the optimal delay values are obtained using the particle swarm optimization algorithm and normalized mutual information criterion. Furthermore, the result is considered as a binary image. Then, 19 features are extracted from the image and the results are presented in the multilayer perceptron neural network for the purpose of training and testing.

Results:

In order to classify N-VF, VT-SVT, N-SVT, VF-VT, VT-N-VF, N-SVT-VF, VT-VF-SVT and VT-VF-SVT-N in the conducted experiments, the accuracy rates were determined at 99.5%, 100%, 94.98%, 100%,100%, 100%, 99.5%, 96.5% and 95%, respectively.

Conclusion:

In this paper, a new approach was developed to classify the abnormal signals obtained from an ECG such as VT, VF, and SVT compared to a normal signal. Compared to Other related studies, our proposed system significantly performed better

ARTIFICIAL NEURAL NETWORK BASED ECG ARRHYTHMIA CLASSIFICATION

Reliable and computationally efficient means of classifying electrocardiogram (ECG) signals has been the subject of considerable research effort in recent years. This paper explores the potential applications of a talented, versatile computation model called the Artificial Neural Network (ANN) in the field of ECG signal classification. Two types of ANNs: Multi-Layered Feed Forward Network (MLFFN) and Probabilistic Neural Networks (PNN) are used to classify seven types of ECG beats. It includes six types of arrhythmia data and normal data. Here, parametric modeling strategies are used in conjunction with ANN classifiers to discriminate ECG signals. Instead of giving the ECG data as such, parameters such as fourth order Auto Regressive model coefficients and Spectral Entropy of the signals has been selected. On testing with the Massachusetts Institute of Technology-Beth Israel Hospital (MIT/BIH) arrhythmia database, it has been observed that PNN has better performance than conventionally used MLFFN in ECG arrhythmia classification. MLFFN with Back Propagation Algorithm gives a classification accuracy of 97.54% and PNN gives 98.96%. The classification by PNN also has an advantage that the computation time for classification is lower than that of MLFFN.

Classification of arrhythmia using hybrid networks

Reliable detection of arrhythmias based on digital processing of Electrocardiogram (ECG) signals is vital in providing suitable and timely treatment to a cardiac patient. Due to corruption of ECG signals with multiple frequency noise and presence of multiple arrhythmic events in a cardiac rhythm, computerized interpretation of abnormal ECG rhythms is a challenging task. This paper focuses a Fuzzy C- Mean (FCM) clustered Probabilistic Neural Network (PNN) and Multi Layered Feed Forward Network (MLFFN) for the discrimination of eight types of ECG beats. Parameters such as fourth order Auto Regressive (AR) coefficients along with Spectral Entropy (SE) are extracted from each ECG beat and feature reduction has been carried out using FCM clustering. The cluster centers form the input of neural network classifiers. The extensive analysis of Massachusetts Institute of Technology- Beth Israel Hospital (MIT-BIH) arrhythmia database shows that FCM clustered PNNs is superior in cardiac arrhythmia classification than FCM clustered MLFFN with an overall accuracy of 99.05%, 97.14%, respectively.

Fuzzy clustered probabilistic and multi layered feed forward neural networks for electrocardiogram arrhythmia classification

The role of electrocardiogram (ECG) as a noninvasive technique for detecting and diagnosing cardiac problems cannot be overemphasized. This paper introduces a fuzzy C-mean (FCM) clustered probabilistic neural network (PNN) for the discrimination of eight types of ECG beats. The performance has been compared with FCM clustered multi layered feed forward network (MLFFN) trained with back propagation algorithm. Important parameters are extracted from each ECG beat and feature reduction has been carried out using FCM clustering. The cluster centers form the input of neural network classifiers. The extensive analysis using the MIT-BIH arrhythmia database has shown an average classification accuracy of 97.54% with FCM clustered MLFFN and 99.58% with FCM clustered PNN. Fuzzy clustering improves the classification speed as well. The result reveals the capability of the FCM clustered PNN in the computer-aided diagnosis of ECG abnormalities.

A pilot study examining the performance of polynomial-modeled ventricular shock electrograms for rhythm discrimination in implantable devices

BACKGROUND

Inappropriate shocks continue to be a problem for patients with implantable defibrillators (ICD). We evaluated the performance of polynomial-modeled ventricular electrograms (EGM) to discriminate between supraventricular tachycardia (SVT) and ventricular tachycardia (VT).

METHODS

Seven sets of EGM from patients having both SVT and VT documented during a single ICD interrogation were included. The cardiac cycle was analyzed off-line in two parts, QR and RQ segments, which were modeled separately using third-order and sixth-order polynomial equations, respectively. These segments were then analyzed to determine which polynomial coefficients were most significant for rhythm discrimination.

RESULTS

When analyzing the QR segment during arrhythmia, there were statistically significant (P<0.05) correlations in 4 of 4 (100%) of the QR coefficients when comparing normal sinus rhythm (NSR) to SVT and 2 of 4 (50%) when comparing NSR to VT or SVT to VT. When analyzing the RQ segment during arrhythmia, there were statistically significant (P<0.05) correlations in 4 of 7 (57%) of the RQ coefficients when comparing NSR to SVT, 5 of 7 (71%) when comparing NSR to VT, and 3 of 7 (43%) when comparing SVT to VT. Using a cutoff value of 50% change from NSR, the ratio of first-order to zero-order QR coefficient was able to completely separate VT from SVT (P=0.03) in this series of patients.

CONCLUSION

Our data demonstrate the feasibility of simple polynomial equations that reproduce the depolarization and repolarization phases of human ventricular shock EGM. The ratio of first-order to zero-order QR coefficient was able to reliably discriminate between SVT and VT while reducing the polynomial model to a first-order system. The results of this pilot trial may serve as the basis for a larger prospective trial implementing a discrimination algorithm for use in low computational power implantable devices.

A segmental polynomial model of ventricular electrograms as a simple and efficient morphology discriminator for implantable devices

BACKGROUND

The goal of this study is to construct a polynomial model of the ventricular electrogram (EGM) that faithfully reproduces the EGM and can be implemented in current, low computational power implantable devices. Such a model of ventricular EGMs is still lacking.

METHODS

New Zealand White rabbits underwent chronic implantation of pacemakers through a left thoracotomy approach. Unipolar ventricular EGMs sampled at a frequency of 1 kHz were stored digitally in 1-minute segments before and after intravenous injection of isoproterenol or procainamide. Each cardiac cycle was divided into a QR and an RQ segment which were modeled separately using a 6th order polynomial equation.

RESULTS

The 14 coefficients of each cardiac cycle were reproducible throughout the baseline recordings (r > or = 0.94, P < 0.002). Isoproterenol caused no changes in the coefficients of the QR segment but significantly altered all but one of the seven coefficients of the RQ segment (p(6)= 0.0039, p(5)= 0.017, p(4)= 0.00007, p(3)= 0.112, p(2)= 0.00016, p(1)= 0.0086, p(a)= 0.00003). Procainamide caused statistically significant changes in both QR segment (p(6)= 0.018, p(5)= 0.287, p(4)= 0.019, p(3)= 0.176, p(2)= 0.016, p(1)= 0.362, p(a)= 0.000044) and RQ segment (p(6)= 0.0028, p(5)= 0.036, p(4)= 0.002, p(3)= 0.058, p(2)= 0.022, p(1)= 0.718, p(a)= 0.0018) coefficients.

CONCLUSION

Our data demonstrate the feasibility of a segmental polynomial equation that reproduces the phases of depolarization and repolarization of the rabbit EGM. This model is reproducible and demonstrates the expected changes with antiarrhythmic drug administration. If reproduced in humans, these findings can have wide applications in patients with implantable devices, ranging from morphologic discrimination of arrhythmias to early detection of metabolic derangements or drug effects.

Cardiac arrhythmia classification using autoregressive modeling

BACKGROUND

Computer-assisted arrhythmia recognition is critical for the management of cardiac disorders. Various techniques have been utilized to classify arrhythmias. Generally, these techniques classify two or three arrhythmias or have significantly large processing times. A simpler autoregressive modeling (AR) technique is proposed to classify normal sinus rhythm (NSR) and various cardiac arrhythmias including atrial premature contraction (APC), premature ventricular contraction (PVC), superventricular tachycardia (SVT), ventricular tachycardia (VT) and ventricular fibrillation (VF).

METHODS

AR Modeling was performed on ECG data from normal sinus rhythm as well as various arrhythmias. The AR coefficients were computed using Burg's algorithm. The AR coefficients were classified using a generalized linear model (GLM) based algorithm in various stages.

RESULTS

AR modeling results showed that an order of four was sufficient for modeling the ECG signals. The accuracy of detecting NSR, APC, PVC, SVT, VT and VF were 93.2% to 100% using the GLM based classification algorithm.

CONCLUSION

The results show that AR modeling is useful for the classification of cardiac arrhythmias, with reasonably high accuracies. Further validation of the proposed technique will yield acceptable results for clinical implementation.

Use of approximate entropy measurements to classify ventricular tachycardia and fibrillation

Implantable cardiac devices which treat arrhythmias require automated detection to decide when to deliver therapy and, in some devices like the implantable cardioverter-defibrillator (ICD), to determine which therapy to deliver. This type of detection often requires the separation of fibrillatory chaotic rhythms from coherent tachycardias. In the ICD, multiple rate zones can be programmed for detection of ventricular tachycardia (VT) and ventricular fibrillation (VF) and for delivery of therapy: antitachycardia pacing and cardioversion for VT, and defibrillation for VF. Previous research determined that current technology is unable to uniquely classify VF. Analysis of typical settings of the ICD revealed that 40% to 80% of VTs were misclassified as VF (depending on device/setting). Improved detection of VF must be sought in order to capitalize on the cost effectiveness of low-energy VT therapies, potentially saving up to 20% of the life of the battery. In this study, a statistical measure of variability, called approximate entropy (ApEn), has been applied to separate fibrillatory and nonfibrillatory rhythms. Although standard deviation is often used as a measure of variability, it fails to capture the level of regularity or complexity. ApEn improves upon standard deviation by quantifying differences between random and regular signals. ApEn was tested on a small patient set containing VF, VT, and sinus rhythm (SR). Intracardiac recordings (bipolar, ventricular) were selected from the Ann Arbor Electrogram Library (Ann Arbor, MI) where filtering (1-500 Hz), amplification (approximately 1000), and digitization (1000 Hz) are tightly controlled. Results demonstrated that ApEn has the ability to quantify subtle differences between VF and other rhythms.

Comprehensive scheme for detection of ventricular fibrillation for implantable cardioverter defibrillators

Implantable cardioverter defibrillators (ICDs) detect and defibrillate ventricular fibrillation (VF) and ventricular tachycardia (VT). Other therapies which use less energy are also available to terminate VT. Previous studies have shown that ICD rate schemes often misdiagnose VT as VF. In this study, an improved VF classification scheme was designed and tested, which employs the classic rate criteria plus paired signal concordance (PSC); PSC uniquely detects VF where VT and VF rates overlap (220-340 ms). Two signals from a bipolar pair (1 cm) recorded in a unipolar sense exhibit similar signal shape for concordant rhythms, such as sinus rhythm and VT, and disconcordance for VF. Once the rate criterion is met, PSC is measured by the peak normalized cross-correlation coefficient calculated over the depolarization. Variability, measured by a modified range, determined the contextual diagnosis over a passage. Sinus rhythm (20), VT (12), VF (22), atrial fibrillation (10), sinus rhythm with ventricular premature depolarizations (7), and polymorphic VT (4) passages were recorded from 38 patients. Rate-PSC was tested with unfiltered, digitized signals (1-500 Hz, 1,000 samples per second) and with filtered, downsampled signals (1-50 Hz, 100 samples per second). Sensitivity values, or percentage of correct VF detection, and specificity values, or detection of all other rhythms, were generated and compared with simulations of three commercial ICDs programmed to similar settings as rate-PSC and to nominal settings. The sensitivity values for rate-PSC with unfiltered and with filtered signals and for ICDs with 220 ms and with nominal settings were 100%, 100%, 48-80%, and 100%, respectively; the corresponding specificity values were 95%, 83%, 93%, and 7-13%, respectively. It was concluded that the rate-PSC scheme was able to reliably separate VF from other rhythms, even rhythms that have a variable morphology or variable rate. With the confidence of accurate VF detection, use of low-energy therapies for non-VF rhythms will increase device longevity and enhance patient comfort.