ECG Arrhythmia Detection and Classification Using Relevance Vector Machine

The aim of this paper is twofold. First, we present a thorough experimental study to show the superiority of the generalization capability of the support vector machine (SVM) approach in the automatic classification of electrocardiogram (ECG) beats. Second, we propose a novel classification system based on particle swarm optimization (PSO) to improve the generalization performance of the SVM classifier. For this purpose, we have optimized the SVM classifier design by searching for the best value of the parameters that tune its discriminant function, and upstream by looking for the best subset of features that feed the classifier. The experiments were conducted on the basis of ECG data from the Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) arrhythmia database to classify five kinds of abnormal waveforms and normal beats. In particular, they were organized so as to test the sensitivity of the SVM classifier and that of two reference classifiers used for comparison, i.e., the k-nearest neighbor (kNN) classifier and the radial basis function (RBF) neural network classifier, with respect to the curse of dimensionality and the number of available training beats. The obtained results clearly confirm the superiority of the SVM approach as compared to traditional classifiers, and suggest that further substantial improvements in terms of classification accuracy can be achieved by the proposed PSO-SVM classification system. On an average, over three experiments making use of a different total number of training beats (250, 500, and 750, respectively), the PSO-SVM yielded an overall accuracy of 89.72% on 40438 test beats selected from 20 patient records against 85.98%, 83.70%, and 82.34% for the SVM, the kNN, and the RBF classifiers, respectively.

Fault diagnosis methods based on machine learning have made great progress for rotating machinery. The main steps of the machine learning process involve feature extraction, selection, and classification. Feature selection improves classification accuracy and reduces diagnosis time by selecting the better features. Due to the difficulty of traditional feature selection methods to rank the feature importance of each class, the best subset of features could hardly be obtained. Therefore, this article proposes a new feature selection method to address the shortcomings of the above traditional methods, called Feature Ranking based on Optimal Class Distance Ratio (FROCDR), which can choose the optimal features between every two classes of samples to obtain feature ranking that is conducive to classification. In order to comprehensively extract the fault information in the signal, the multiscale analysis and the variational mode decomposition (VMD) method are applied to process the vibration signals under different scales and frequency bands, and the processed signals are visualized by symmetrized dot pattern (SDP). In addition, features are extracted from the obtained SDP images, and the proposed FROCDR method is used to select the best subset of features. The final diagnosis task is accomplished by a random forest (RF) classifier. Experimental cases of bearing and gear data show that the proposed method has higher diagnostic accuracy and stability.

Due to the high sampling rate, the recorded Electrocardiograms (ECG) data are huge. For storing and transmitting ECG data, wide spaces and more bandwidth are therefore needed. The ECG data are also very important to preprocessing and compress so that it is distributed and processed with less bandwidth and less space effectively. This manuscript is aimed at creating an effective ECG compression method. The reported ECG data are processed first in the pre-processing unit (ProUnit) in this method. In this unit, ECG data have been standardized and segmented. The resulting ECG data would then be sent to the Compression Unit (CompUnit). The unit consists of an algorithm for lossy compression (LosyComp), with a lossless algorithm for compression (LossComp). The randomness ECG data is transformed into high randomness data by the failure compression algorithm. The data's high redundancy is then used with the LosyComp algorithm to reach a high compression ratio (CR) with no degradation. The LossComp algorithms recommended in this manuscript are the Discrete Cosine Transform (DCT) and Discrete Wavelet Transform (DWT). LossComp algorithms such as Arithmetic Encoding (Arithm) and Run Length Encoding (RLE) are also suggested. To evaluate the proposed method, we measure the Compression Time (CompTime), and Reconstruction Time (RecTime) (T), RMSE and CR. Simulation results suggest the highest output in compression ratio and in complexity by adding RLE after the DCT algorithm. The simulation findings indicate that the inclusion of RLE following the DCT algorithm increases performance in terms of CR and complexity. With CR = 55% with RMSE = 0:14 and above 94% with RMSE = 0:2, DCT as a LossComp algorithm was utilized initially, followed by RLE as a LossComp algorithm.

Application of the PSO–SVM model for recognition of control chart patterns

To determine viability of drowsiness detection, researchers study the feasibility of photoplethysmogram (PPG) data collection from the geography of the aviation headset, correlating to electrocardiogram (ECG) reference. Fatigue has been a probable cause, contributing factor, or a finding in 20% of transportation incidents and accidents studied between January 2001 and December 2012. This operational hazard is particularly troublesome within aviation and airline operations. PPG and ECG data were collected synchronously from Federal Aviation Administration (FAA) commercially rated pilots during flight simulation in the window of circadian low (WOCL). Valid PPG and ECG data from 14 participants were analyzed, which yielded approximately 2 hr of data per participant for fatigue-related analysis. The results of the study demonstrate clear trends toward decreased heart rate for both ECG and PPG and suggest progression of drowsiness between four separate periods (T1, T2, T3, and T4) selected during the study; however, the mean heart rate change from T1 to T4 was statistically significant. The results suggest that ECG and PPG data can be an important tool to observe conditions where drowsiness or fatigue may add risk to the operation. In addition, the data show high correlation between ECG and PPG data, further suggesting that a simpler PPG sensor, mounted within the geography of the aviation headset, may streamline the operationalization of important physiological data. Incorporation of PPG sensors and associated signal processing methods into facilitating equipment, such as the aviation headset, may add a layer to operational safety.

It is of great importance to extract and validate an optimal subset of non-dominated features for effective multi-label classification. However, deciding on the best subset of features is an NP-Hard problem and plays a key role in improving the prediction accuracy and the processing time of video datasets. In this study, we propose autoencoders for dimensionality reduction of video data sets and ensemble the features extracted by the multi-objective evolutionary Non-dominated Sorting Genetic Algorithm and the autoencoder. We explore the performance of well-known multi-label classification algorithms for video datasets in terms of prediction accuracy and the number of features used. More specifically, we evaluate Non-dominated Sorting Genetic Algorithm-II, autoencoders, ensemble learning algorithms, Principal Component Analysis, Information Gain, and Correlation Based Feature Selection. Some of these algorithms use feature selection techniques to improve the accuracy of the classification. Experiments are carried out with local feature descriptors extracted from two multi-label datasets, the MIR-Flickr dataset which consists of images and the Wireless Multimedia Sensor dataset that we have generated from our video recordings. Significant improvements in the accuracy performance of the algorithms are observed while the number of features is being reduced.

In the past decade, many studies have been conducted to advance computer-aided systems for Alzheimer’s disease (AD) diagnosis. Most of them have recently developed systems concentrated on extracting and combining features from MRI, PET, and CSF. For the most part, they have obtained very high performance. However, improving the performance of a classification problem is complicated, specifically when the model’s accuracy or other performance measurements are higher than 90%. In this study, a novel methodology is proposed to address this problem, specifically in Alzheimer’s disease diagnosis classification. This methodology is the first of its kind in the literature, based on the notion of replication on the feature space instead of the traditional sample space. Briefly, the main steps of the proposed method include extracting, embedding, and exploring the best subset of features. For feature extraction, we adopt VBM-SPM; for embedding features, a concatenation strategy is used on the features to ultimately create one feature vector for each subject. Principal component analysis is applied to extract new features, forming a low-dimensional compact space. A novel process is applied by replicating selected components, assessing the classification model, and repeating the replication until performance divergence or convergence. The proposed method aims to explore most significant features and highest-preforming model at the same time, to classify normal subjects from AD and mild cognitive impairment (MCI) patients. In each epoch, a small subset of candidate features is assessed by support vector machine (SVM) classifier. This repeating procedure is continued until the highest performance is achieved. Experimental results reveal the highest performance reported in the literature for this specific classification problem. We obtained a model with accuracies of 98.81%, 81.61%, and 81.40% for AD vs. normal control (NC), MCI vs. NC, and AD vs. MCI classification, respectively.

Graphs help to define the relationships between entities in the data. These relationships, represented by edges, often provide additional context information which can be utilised to discover patterns in the data. Graph Neural Networks (GNNs) employ the inductive bias of the graph structure to learn and predict on various tasks. The primary operation of graph neural networks is the feature aggregation step performed over neighbours of the node based on the structure of the graph. In addition to its own features, for each hop, the node gets additional combined features from its neighbours. These aggregated features help define the similarity or dissimilarity of the nodes with respect to the labels and are useful for tasks like node classification. However, in real‐world data, features of neighbours at different hops may not correlate with the node's features. Thus, any indiscriminate feature aggregation by GNN might cause the addition of noisy features leading to degradation in model's performance. In this work, we show that selective aggregation of node features from various hops leads to better performance than default aggregation on the node classification task. Furthermore, we propose a Dual‐Net GNN architecture with a classifier model and a selector model. The classifier model trains over a subset of input node features to predict node labels while the selector model learns to provide optimal input subset to the classifier for the best performance. These two models are trained jointly to learn the best subset of features that give higher accuracy in node label predictions. With extensive experiments, we show that our proposed model outperforms both feature selection methods and state‐of‐the‐art GNN models with remarkable improvements up to 27.8%.

An electrocardiogram (ECG) measures the heart's electrical function and has been widely used to detect heart disorders because of their simplicity and non-obtrusive nature. Nowadays, in medical decision support systems, soft computing methods are commonly used. Medical diagnosis helps to acquire various characteristics reflecting the disease's various variations. It is likely to have important, irrelevant, and redundant features to reflect disease, with the aid of various diagnostic procedures. It is a non-trivial job to define a good feature subset for efficient classification. The objective of this research is to assess three different soft computing-based unsupervised feature reduction approaches namely Soft Set-Based Unsupervised Feature Reduction, Unsupervised Quick Reduction, and Unsupervised Relative Reduction, to evaluate the best subset of features with enhanced diagnostic classification accuracy for cardiovascular disease. The output of feature selection algorithms is evaluated with the heart disease dataset from the UCI repository. The experimental findings indicate that the Soft Set-based Unsupervised Feature Redact algorithm achieved good classification accuracy compared to other feature reduction approaches.KeywordsECG signalClassificationFeature selectionSoft computing TechniquesSoft set theoryRough set theory

R peak detection is crucial in electrocardiogram (ECG) signal analysis to detect and diagnose cardiovascular diseases (CVDs). Herein, the dynamic mode selected energy (DMSE) and adaptive window sizing (AWS) algorithm are proposed for detecting R peaks with better efficiency. The DMSE algorithm adaptively separates the QRS components and all non-objective components from the ECG signal. Based on local peaks in QRS components, the AWS algorithm adaptively determines the Region of Interest (ROI). The Feature Extraction process computes the statistical properties of energy, frequency, and noise from each ROI. The Sequential Forward Selection (SFS) procedure is used to find the best subsets of features. Based on these characteristics, an ensemble of decision tree algorithms detects the R peaks. Finally, the R peak position on the initial ECG signal is adjusted using the R location correction (RLC) algorithm. The proposed method has an experimental accuracy of 99.94%, a sensitivity of 99.98%, positive predictability of 99.96%, and a detection error rate of 0.06%. Given the high efficiency in detection and fast processing speed, the proposed approach is ideal for intelligent medical and wearable devices in the diagnosis of CVDs.

Simultaneous Monitoring of ECG and EDA Using a Wearable Armband for Analyzing Sympathetic Nerve Activity

Objective: To measure prevalence of discordance between electrical activity recorded by electrocardiography (ECG) and myocardial activity visualized by echocardiography (echo) in patients presenting after cardiac arrest and to compare survival outcomes in cohorts defined by ECG and echo. Methods: This is a secondary analysis of a previously published prospective study at twenty hospitals. Patients presenting after out-of-hospital arrest were included. The cardiac electrical activity was defined by ECG and contemporaneous myocardial activity was defined by bedside echo. Myocardial activity by echo was classified as myocardial asystole- -the absence of myocardial movement, pulseless myocardial activity (PMA)--visible myocardial movement but no pulse, and myocardial fibrillation- -visualized fibrillation. Primary outcome was the prevalence of discordance between electrical activity and myocardial activity. Secondary outcome was survival to hospital discharge. Results: 793 patients and 1943 pauses in CPR were included. 28.6% of CPR pauses demonstrated a difference in electrical activity (ECG) and myocardial activity (echo), 5.0% with asystole (ECG) and PMA (echo), and 22.1% with PEA (ECG) and myocardial asystole (echo). Survival to hospital admission for patients with PMA (echo) was 29.1% (95%CI-23.9-34.9) compared to those with PEA (ECG) (21.4%, 95%CI-17.7-25.6). Twenty-five percent of the 32 pauses in CPR with a shockable rhythm by echo demonstrated a non-shockable rhythm by ECG and were not defibrillated. One of these patients survived, a patient with asystole on ECG and vfib by echo survived because vfib was identified on ECG during a subsequent pause and was defibrillated. Conclusion: Patients in cardiac arrest commonly demonstrate different electrical (ECG) and myocardial activity (echo). Further research is needed to better define cardiac activity during cardiac arrest and to explore outcome between groups defined by electrical and myocardial activity.

Owing to a lack of easy access to an individual's electrocardiogram (ECG) data that may be distributed across multiple hospitals, medical practice using serial ECG analysis has not so far been extensively used. Motivated to address this problem, we design and implement an architecture for cross-hospital access to ECG data. The architecture provides a Web portal for retrieving a particular patient's ECG data distributed across a number of ECG data sources. Two important features of this architecture are (1) the distributed search mechanism that uses a master patient index to determine what ECG data sources are likely to contain the ECG data pertaining to a specified patient and (2) the Web service-based interoperable method for accessing disparate ECG data sources. We show that the architecture provides an effective search and retrieval facility for cross-hospital access to ECG data; we show this by implementing and employing a prototype in an experimental scenario consisting of two disparate ECG data sources located at two sites. Our study demonstrates a significant effort to improve the ease of cross-hospital access to ECG data.

Comparison of outcomes between pulseless electrical activity by electrocardiography and pulseless myocardial activity by echocardiography in out-of-hospital cardiac arrest; secondary analysis from a large, prospective study

The purpose of this study is to present a hybrid approach based on the combined use of a genetic algorithm (GA) and a nearest neighbours classifier for the selection of the critical clinical features which are strongly related with the incidence of fatal and non fatal Cardiovascular Disease (CVD) in patients with Type 2 Diabetes Mellitus (T2DM). For the development and the evaluation of the proposed algorithm, data from the medical records of 560 patients with T2DM are used. The best subsets of features proposed by the implemented algorithm include the most common risk factors, such as age at diagnosis, duration of diagnosed diabetes, glycosylated haemoglobin (HbA1c), cholesterol concentration, and smoking habit, but also factors related to the presence of other diabetes complications and the use of antihypertensive and diabetes treatment drugs (i.e. proteinuria, calcium antagonists, b-blockers, diguanides and insulin). The obtained results demonstrate that the best performance was achieved when the weighted k-nearest neighbours classifier was applied to the CVD dataset with the best subset of features selected by the GA, which resulted in high levels of accuracy (0.96), sensitivity (0.80) and specificity (0.98).