SpEC: A system for patient specific ECG beat classification using deep residual network

A pilot study on AI-driven approaches for classification of mental health disorders

The increasing prevalence of mental disorders among youth worldwide is one of society's most pressing issues. The proposed methodology introduces an artificial intelligence-based approach for comprehending and analyzing the prevalence of neurological disorders. This work draws upon the analysis of the Cities Health Initiative dataset. It employs advanced machine learning and deep learning techniques, integrated with data science, statistics, optimization, and mathematical modeling, to correlate various lifestyle and environmental factors with the incidence of these mental disorders. In this work, a variety of machine learning and deep learning models with hyper-parameter tuning are utilized to forecast trends in the occurrence of mental disorders about lifestyle choices such as smoking and alcohol consumption, as well as environmental factors like air and noise pollution. Among these models, the convolutional neural network (CNN) architecture, termed as DNN1 in this paper, accurately predicts mental health occurrences relative to the population mean with a maximum accuracy of 99.79%. Among the machine learning models, the XGBoost technique yields an accuracy of 95.30%, with an area under the ROC curve of 0.9985, indicating robust training. The research also involves extracting feature importance scores for the XGBoost classifier, with Stroop test performance results attaining the highest importance score of 0.135. Attributes related to addiction, namely smoking and alcohol consumption, hold importance scores of 0.0273 and 0.0212, respectively. Statistical tests on the training models reveal that XGBoost performs best on the mean squared error and R-squared tests, achieving scores of 0.013356 and 0.946481, respectively. These statistical evaluations bolster the models' credibility and affirm the best-fit models' accuracy. The proposed research in the domains of mental health, addiction, and pollution stands to aid healthcare professionals in diagnosing and treating neurological disorders in both youth and adults promptly through the use of predictive models. Furthermore, it aims to provide valuable insights for policymakers in formulating new regulations on pollution and addiction.

ECG data enhancement method using generate adversarial networks based on Bi-LSTM and CBAM

Objective.The classification performance of electrocardiogram (ECG) classification algorithms is easily affected by data imbalance, which often leads to poor model prediction performance for a few classes and a consequent decrease in the overall performance of the model.Approach.To address this problem, this paper proposed an ECG data augmentation method based on a generative adversarial network (GAN) that combines bidirectional long short-term memory (Bi-LSTM) networks and convolutional block attention mechanism (CBAM) to improve the overall performance of ECG classification models. In this paper, we used two ECG databases, namely the MIT-BIH arrhythmia (MIT-BIH-AR) database and the Chinese cardiovascular disease database (CCDD). The quality of the ECG signals produced by the generated models was assessed using the percent relative difference, root mean square error, Frechet distance, dynamic time warping (DTW), and Pearson correlation metrics. In addition, we also validated the impact of our proposed data augmentation method on ECG classification performance on MIT-BIH-AR database and CCDD.Main results.On the MIT-BIH-AR database, the overall accuracy of the data-enhanced balanced dataset was improved to 99.46% for 15 types of heartbeat classification task. On the CCDD, which focuses on the detection of ventricular precession (PVC), the overall accuracy of PVC detection improved to 99.15% after performing data enhancement.Significance.The experimental results indicate that the data augmentation method proposed in this paper can further improve the ECG classification performance.

Wireless CardioS framework for continuous ECG acquisition

A first-level textile-based electrocardiogram (ECG) monitoring system referred to as "CardioS" (cardiac sensor) for continuous health monitoring applications is proposed in this study to address the demand for resource-constrained environments. and the signal quality assessment of a wireless CardioS was studied. The CardioS consists of a Lead-I ECG signal recorded wirelessly using silver-plated nylon woven (Ag-NyW) dry textile electrodes to compare the results of wired wearable Ag-NyW textile electrode-based ECG acquisition system and CardioS. The effect of prolonged usage of Ag-NyW dry electrodes on electrode impedance was tested in the current work. In addition, electrode half-cell potential was measured to validate the range of Ag-NyW dry electrodes for ECG signal acquisition. Further, the quality of signals recorded by the proposed wireless CardioS framework was evaluated and compared with clinical disposable (Ag-AgCl Gel) electrodes. The signal quality was assessed in terms of mean magnitude coherence spectra, signal cross-correlation, signal-to-noise-band ratio (Sband/Nband), crest factor, low and high band powers and power spectral density. The experimental results showed that the impedance was increased by 2.5-54.6% after six weeks of continuous usage. This increased impedance was less than 1 MΩ/cm2, as reported in the literature. The half-cell potential of the Ag-NyW textile electrode obtained was 80 mV, sufficient to acquire the ECG signal from the human body. All the fidelity parameters measured by Ag-NyW textile electrodes were correlated with standard disposable electrodes. The cardiologists validated all the measurements and confirmed that the proposed framework exhibited good performance for ECG signal acquisition from the five healthy subjects. As a result of its low-cost architecture, the proposed CardioS framework can be used in resource-constrained environments for ECG monitoring.

Cancer detection for small-size and ambiguous tumors based on semantic FPN and transformer

Early detection of tumors has great significance for formative detection and determination of treatment plans. However, cancer detection remains a challenging task due to the interference of diseased tissue, the diversity of mass scales, and the ambiguity of tumor boundaries. It is difficult to extract the features of small-sized tumors and tumor boundaries, so semantic information of high-level feature maps is needed to enrich the regional features and local attention features of tumors. To solve the problems of small tumor objects and lack of contextual features, this paper proposes a novel Semantic Pyramid Network with a Transformer Self-attention, named SPN-TS, for tumor detection. Specifically, the paper first designs a new Feature Pyramid Network in the feature extraction stage. It changes the traditional cross-layer connection scheme and focuses on enriching the features of small-sized tumor regions. Then, we introduce the transformer attention mechanism into the framework to learn the local feature of tumor boundaries. Extensive experimental evaluations were performed on the publicly available CBIS-DDSM dataset, which is a Curated Breast Imaging Subset of the Digital Database for Screening Mammography. The proposed method achieved better performance in these models, achieving 93.26% sensitivity, 95.26% specificity, 96.78% accuracy, and 87.27% Matthews Correlation Coefficient (MCC) value, respectively. The method can achieve the best detection performance by effectively solving the difficulties of small objects and boundaries ambiguity. The algorithm can further promote the detection of other diseases in the future, and also provide algorithmic references for the general object detection field.

An Automatic Premature Ventricular Contraction Recognition System Based on Imbalanced Dataset and Pre-Trained Residual Network Using Transfer Learning on ECG Signal

The development of automatic monitoring and diagnosis systems for cardiac patients over the internet has been facilitated by recent advancements in wearable sensor devices from electrocardiographs (ECGs), which need the use of patient-specific approaches. Premature ventricular contraction (PVC) is a common chronic cardiovascular disease that can cause conditions that are potentially fatal. Therefore, for the diagnosis of likely heart failure, precise PVC detection from ECGs is crucial. In the clinical settings, cardiologists typically employ long-term ECGs as a tool to identify PVCs, where a cardiologist must put in a lot of time and effort to appropriately assess the long-term ECGs which is time consuming and cumbersome. By addressing these issues, we have investigated a deep learning method with a pre-trained deep residual network, ResNet-18, to identify PVCs automatically using transfer learning mechanism. Herein, features are extracted by the inner layers of the network automatically compared to hand-crafted feature extraction methods. Transfer learning mechanism handles the difficulties of required large volume of training data for a deep model. The pre-trained model is evaluated on the Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) Arrhythmia and Institute of Cardiological Technics (INCART) datasets. First, we used the Pan-Tompkins algorithm to segment 44,103 normal and 6423 PVC beats, as well as 106,239 normal and 9987 PVC beats from the MIT-BIH Arrhythmia and IN-CART datasets, respectively. The pre-trained model employed the segmented beats as input after being converted into 2D (two-dimensional) images. The method is optimized with the using of weighted random samples, on-the-fly augmentation, Adam optimizer, and call back feature. The results from the proposed method demonstrate the satisfactory findings without the using of any complex pre-processing and feature extraction technique as well as design complexity of model. Using LOSOCV (leave one subject out cross-validation), the received accuracies on MIT-BIH and INCART are 99.93% and 99.77%, respectively, suppressing the state-of-the-art methods for PVC recognition on unseen data. This demonstrates the efficacy and generalizability of the proposed method on the imbalanced datasets. Due to the absence of device-specific (patient-specific) information at the evaluating stage on the target datasets in this study, the method might be used as a general approach to handle the situations in which ECG signals are obtained from different patients utilizing a variety of smart sensor devices.

A Reparameterization Multifeature Fusion CNN for Arrhythmia Heartbeats Classification

Aiming at arrhythmia heartbeats classification, a novel multifeature fusion deep learning-based method is proposed. The stationary wavelet transforms (SWT) and RR interval features are firstly extracted. Based on the traditional one-dimensional convolutional neural network (1D-CNN), a parallel multibranch convolutional network is designed for training. The subband of SWT is input into the multiscale 1D-CNN separately. The output fused with RR interval features are fed to the fully connected layer for classification. To achieve the lightweight network while maintaining the powerful inference capability of the multibranch structure, the redundant branches of the network are removed by reparameterization. Experimental results and analysis show that it outperforms existing methods by many in arrhythmic heartbeat classification.

Electrocardiogram classification using TSST-based spectrogram and ConViT

As an important auxiliary tool of arrhythmia diagnosis, Electrocardiogram (ECG) is frequently utilized to detect a variety of cardiovascular diseases caused by arrhythmia, such as cardiac mechanical infarction. In the past few years, the classification of ECG has always been a challenging problem. This paper presents a novel deep learning model called convolutional vision transformer (ConViT), which combines vision transformer (ViT) with convolutional neural network (CNN), for ECG arrhythmia classification, in which the unique soft convolutional inductive bias of gated positional self-attention (GPSA) layers integrates the superiorities of attention mechanism and convolutional architecture. Moreover, the time-reassigned synchrosqueezing transform (TSST), a newly developed time-frequency analysis (TFA) method where the time-frequency coefficients are reassigned in the time direction, is employed to sharpen pulse traits for feature extraction. Aiming at the class imbalance phenomena in the traditional ECG database, the smote algorithm and focal loss (FL) are used for data augmentation and minority-class weighting, respectively. The experiment using MIT-BIH arrhythmia database indicates that the overall accuracy of the proposed model is as high as 99.5%. Furthermore, the specificity (Spe), F1-Score and positive Matthews Correlation Coefficient (MCC) of supra ventricular ectopic beat (S) and ventricular ectopic beat (V) are all more than 94%. These results demonstrate that the proposed method is superior to most of the existing methods.

Classification of ECG signal using FFT based improved Alexnet classifier

Electrocardiograms (ECG) are extensively used for the diagnosis of cardiac arrhythmias. This paper investigates the use of machine learning classification algorithms for ECG analysis and arrhythmia detection. This is a crucial component of a conventional electronic health system, and it frequently necessitates ECG signal reduction for long-term data storage and remote transmission. Signal processing methods must be used to extract the function of the morphological properties of the ECG signal changing with time, which is difficult to discern in the typical visual depiction of the ECG signal. In biomedical research, signal processing and data analysis are commonly employed methodologies. This work proposes the use of an ECG arrhythmia classification method based on Fast Fourier Transform (FFT) for feature extraction and an improved AlexNet classifier to distinguish the difference between four types of arrhythmia conditions that were collected from records. The Convolutional Neural Network (CNN) algorithm’s results are compared to those of other algorithms, and the simulation results prove that the proposed technique is more effective for various parameters. The final results of the proposed system show that its ability to find deviations is 20% better than that of traditional systems.

Real-Time Patient-Specific ECG Classification by 1D Self-Operational Neural Networks

Despite the proliferation of numerous deep learning methods proposed for generic ECG classification and arrhythmia detection, compact systems with the real-time ability and high accuracy for classifying patient-specific ECG are still few. Particularly, the scarcity of patient-specific data poses an ultimate challenge to any classifier. Recently, compact 1D Convolutional Neural Networks (CNNs) have achieved the state-of-the-art performance level for the accurate classification of ventricular and supraventricular ectopic beats. However, several studies have demonstrated the fact that the learning performance of the conventional CNNs is limited because they are homogenous networks with a basic (linear) neuron model. In order to address this deficiency and further boost the patient-specific ECG classification performance, in this study, we propose 1D Self-organized Operational Neural Networks (1D Self-ONNs). Due to its self-organization capability, Self-ONNs have the utmost advantage and superiority over conventional ONNs where the prior operator search within the operator set library to find the best possible set of operators is entirely avoided. As the first study where 1D Self-ONNs are ever proposed for a classification task, our results over the MIT-BIH arrhythmia benchmark database demonstrate that 1D Self-ONNs can surpass 1D CNNs with a significant margin while having a similar computational complexity. Under AAMI recommendations and with minimal common training data used, over the entire MIT-BIH dataset 1D Self-ONNs have achieved 98% and 99.04% average accuracies, 76.6% and 93.7% average F1 scores on supra-ventricular and ventricular ectopic beat (VEB) classifications, respectively, which is the highest performance level ever reported.

ARRHYTHMIA CLASSIFICATION USING DEEP RESIDUAL NEURAL NETWORKS

Arrhythmia classification with electrocardiogram (ECG) is of great importance for the identification of arrhythmia diseases. However, since the variance of ECG signal in wave appears frequently, it is still a very challenging task to obtain a very good classification result. In this paper, an arrhythmia classification with ECG based on deep residual networks is proposed, of which two improved residual blocks are used to combine soft and hard subsampling. With such blocks, the network can well hold spatial information and improve the classification performance with a simple model structure. Experiments on the MIT-BIH arrhythmia database show that the proposed method obtained an average classification accuracy of 99.59% and an average classification specificity 99.63%, which are 0.26% and 0.57% higher than that of the most state-of-art method based on deep learning.