Automated detection of atrial fibrillation using long short-term memory network with RR interval signals

Clinical knowledge-based ECG abnormalities detection using dual-view CNN-Transformer and external attention mechanism

BACKGROUND

Automatic abnormalities detection based on Electrocardiogram (ECG) contributes greatly to early prevention, computer aided diagnosis, and dynamic analysis of cardiovascular diseases. In order to achieve cardiologist-level performance, deep neural networks have been widely utilized to extract abstract feature representations. However, the mechanical stacking of numerous computationally intensive operations makes traditional deep neural networks suffer from inadequate learning, poor interpretability, and high complexity.

METHOD

To address these limitations, a clinical knowledge-based ECG abnormalities detection model using dual-view CNN-Transformer and external attention mechanism is proposed by mimicking the diagnosis of the clinicians. Considering the clinical knowledge that both the detailed waveform changes within a single heartbeat and the global changes throughout the entire recording have complementary roles in abnormalities detection, we presented a dual-view CNN-Transformer to extract and fuse spatial-temporal features from different views. In addition, the locations of the ECG where abnormalities occur provide more information than other areas. Therefore, two external attention mechanisms are designed and added to the corresponding views to help the network learn efficiently.

RESULTS

Experiment results on the 9-class dataset show that the proposed model achieves an average F1-score of 0.854±0.01 with a higher interpretability and a lower complexity, outperforming the state-of-the-art model.

CONCLUSIONS

Combining all these excellent features, this study provides a credible solution for automatic ECG abnormalities detection.

Enhancing ECG signal classification through pre-trained stacked-CNN embeddings: a transfer learning approach

Rapid and accurate electrocardiogram (ECG) signal classification is crucial in high-stakes healthcare settings. However, existing computational models often struggle to balance high performance with computational efficiency. This study introduces an innovative computational framework that combines transfer learning with traditional machine learning to optimize ECG classification. We use a pre-trained Stacked Convolutional Neural Network (SCNN) to generate high-dimensional feature embeddings, which are then evaluated by an array of machine learning classifiers. Our models demonstrate exceptional performance, particularly when utilizing embeddings from SCNNs trained on diverse datasets. This underscores the importance of data diversity in improving classifier discrimination. Notably, Multilayer Perceptrons (MLPs) stand out for their ability to balance computational efficiency with strong performance, achieving test F1-scores of 0.94 and 1.00 in multi-class and binary tasks on the CinC2017 dataset, and 0.85 and 0.99 on the CPSC2018 dataset. Our approach consistently outperforms existing methods, setting new benchmarks in ECG classification. The synergy between deep learning-based feature extraction and traditional machine learning through transfer learning offers a robust, efficient, and adaptable strategy for ECG classification, addressing a critical research gap and laying the groundwork for future advancements in this crucial healthcare field.

A portable household detection system based on the combination of bidirectional LSTM and residual block for automatical arrhythmia detection

OBJECTIVES

Arrhythmia is an important component of cardiovascular disease, and electrocardiogram (ECG) is a method to detect arrhythmia. Arrhythmia detection is often paroxysmal, and ECG signal analysis is time-consuming and expensive. We propose a model and device for convenient monitoring of arrhythmia at any time.

METHODS

This work proposes a model combining residual block and bidirectional long-term short-term memory network (BiLSTM) to detect and classify ECG signals. Residual blocks can extract deep features and avoid performance degradation caused by convolutional networks. Combined with the feature of BiLSTM to strengthen the connection relationship of the local window, it can achieve a better classification and prediction effect.

RESULTS

Model optimization experiments were performed on the MIT-BIH Atrial Fibrillation Database (AFDB) and MIT-BIH Arrhythmia Database (MITDB). The accuracy simulation results on both long and short signal was higher than 99 %. To further demonstrate the applicability of the model, validation experiments were conducted on MIT-BIH Normal Sinus Rhythm Database (NSRDB) and the Long-Term AF Database (LTAFDB) datasets, and the related recognition accuracy were 99.830 and 91.252 %, respectively. Additionally, we proposed a portable household detection system including an ECG and a blood pressure detection module. The detection accuracy was higher than 98 % using the collected data as testing set.

CONCLUSIONS

Hence, we thought our system can be used for practical application.

Automated identification of atrial fibrillation from single-lead ECGs using multi-branching ResNet

Introduction: Atrial fibrillation (AF) is the most common cardiac arrhythmia, which is clinically identified with irregular and rapid heartbeat rhythm. AF puts a patient at risk of forming blood clots, which can eventually lead to heart failure, stroke, or even sudden death. Electrocardiography (ECG), which involves acquiring bioelectrical signals from the body surface to reflect heart activity, is a standard procedure for detecting AF. However, the occurrence of AF is often intermittent, costing a significant amount of time and effort from medical doctors to identify AF episodes. Moreover, human error is inevitable, as even experienced medical professionals can overlook or misinterpret subtle signs of AF. As such, it is of critical importance to develop an advanced analytical model that can automatically interpret ECG signals and provide decision support for AF diagnostics. Methods: In this paper, we propose an innovative deep-learning method for automated AF identification using single-lead ECGs. We first extract time-frequency features from ECG signals using continuous wavelet transform (CWT). Second, the convolutional neural networks enhanced with residual learning (ReNet) are employed as the functional approximator to interpret the time-frequency features extracted by CWT. Third, we propose to incorporate a multi-branching structure into the ResNet to address the issue of class imbalance, where normal ECGs significantly outnumber instances of AF in ECG datasets. Results and Discussion: We evaluate the proposed Multi-branching Resnet with CWT (CWT-MB-Resnet) with two ECG datasets, i.e., PhysioNet/CinC challenge 2017 and ECGs obtained from the University of Oklahoma Health Sciences Center (OUHSC). The proposed CWT-MB-Resnet demonstrates robust prediction performance, achieving an F1 score of 0.8865 for the PhysioNet dataset and 0.7369 for the OUHSC dataset. The experimental results signify the model's superior capability in balancing precision and recall, which is a desired attribute for ensuring reliable medical diagnoses.

Heart rate detection method based on Ballistocardiogram signal of wearable device：Algorithm development and validation

Background

Heart rate, as the four vital signs of human body, is a basic indicator to measure a person's health status. Traditional electrocardiography (ECG) measurement, which is routinely monitored, requires subjects to wear lead electrodes frequently, which undoubtedly places great restrictions on participants' activities during the normal test. At present, the boom of wearable devices has created hope for non-invasive, simple operation and low-cost daily heart rate monitoring, among them, Ballistocardiogram signal (BCG) is an effective heart rate measurement method, but in the actual acquisition process, the robustness of non-invasive vital sign collection is limited. Therefore, it is necessary to develop a method to improve the robustness of heart rate monitoring.

Objective

Therefore, in view of the problem that the accuracy of untethered monitoring heart rate is not high, we propose a method aimed at detecting the heartbeat cycle based on BCG to accurately obtain the beat-to-beat heart rate in the sleep state.

Methods

In this study, we implement an innovative J-wave detection algorithm based on BCG signals. By collecting BCG signals recorded by 28 healthy subjects in different sleeping positions, after preprocessing, the data feature set is formed according to the clustering of morphological features in the heartbeat interval. Finally, a J-wave recognition model is constructed based on bi-directional long short-term memory (BiLSTM), and then the number of J-waves in the input sequence is counted to realize real-time detection of heartbeat. The performance of the proposed heartbeat detection scheme is cross-verified, and the proposed method is compared with the previous wearable device algorithm.

Results

The accuracy of J wave recognition in BCG signal is 99.67%, and the deviation rate of heart rate detection is only 0.27%, which has higher accuracy than previous wearable device algorithms. To assess consistency between method results and heart rates obtained by the ECG, seven subjects are compared using Bland-Altman plots, which show no significant difference between BCG and ECG results for heartbeat cycles.

Conclusions

Compared with other studies, the proposed method is more accurate in J-wave recognition, which improves the accuracy and generalization ability of BCG-based continuous heartbeat cycle extraction, and provides preliminary support for wearable-based untethered daily monitoring.

Few-shot transfer learning for personalized atrial fibrillation detection using patient-based siamese network with single-lead ECG records

The proliferation of wearable devices has allowed the collection of electrocardiogram (ECG) recordings daily to monitor heart rhythm and rate. For example, 24-hour Holter monitors, cardiac patches, and smartwatches are widely used for ECG gathering and application. An automatic atrial fibrillation (AF) detector is required for timely ECG interpretation. Deep learning models can accurately identify AFs if large amounts of annotated data are available for model training. However, it is impractical to request sufficient labels for ECG recordings for an individual patient to train a personalized model. We propose a Siamese-network-based approach for transfer learning to address this issue. A pre-trained Siamese convolutional neural network is created by comparing two labeled ECG segments from the same patient. We sampled 30-second ECG segments with a 50% overlapping window from the ECG recordings of patients in the MIT-BIH Atrial Fibrillation Database. Subsequently, we independently detected the occurrence of AF in each patient in the Long-Term AF Database. By fine-tuning the model with the 1, 3, 5, 7, 9, or 11 ECG segments ranging from 30 to 180 s, our method achieved macro-F1 scores of 96.84%, 96.91%, 96.97%, 97.02%, 97.05%, and 97.07%, respectively.

MCHeart: Multi-Channel-Based Heart Signal Processing Scheme for Heart Noise Detection Using Deep Learning

In this study, we constructed a model to predict abnormal cardiac sounds using a diverse set of auscultation data collected from various auscultation positions. Abnormal heart sounds were identified by extracting features such as peak intervals and noise characteristics during systole and diastole. Instead of using raw signal data, we transformed them into log-mel 2D spectrograms, which were employed as input variables for the CNN model. The advancement of our model involves integrating a deep learning architecture with feature extraction techniques based on existing knowledge of cardiac data. Specifically, we propose a multi-channel-based heart signal processing (MCHeart) scheme, which incorporates our proposed features into the deep learning model. Additionally, we introduce the ReLCNN model by applying residual blocks and MHA mechanisms to the LCNN architecture. By adding murmur features with a smoothing function and training the ReLCNN model, the weighted accuracy of the model increased from 79.6% to 83.6%, showing a performance improvement of approximately 4% point compared to the LCNN baseline model.

Deep learning for ECG Arrhythmia detection and classification: an overview of progress for period 2017-2023

Cardiovascular diseases are a leading cause of mortality globally. Electrocardiography (ECG) still represents the benchmark approach for identifying cardiac irregularities. Automatic detection of abnormalities from the ECG can aid in the early detection, diagnosis, and prevention of cardiovascular diseases. Deep Learning (DL) architectures have been successfully employed for arrhythmia detection and classification and offered superior performance to traditional shallow Machine Learning (ML) approaches. This survey categorizes and compares the DL architectures used in ECG arrhythmia detection from 2017-2023 that have exhibited superior performance. Different DL models such as Convolutional Neural Networks (CNNs), Multilayer Perceptrons (MLPs), Transformers, and Recurrent Neural Networks (RNNs) are reviewed, and a summary of their effectiveness is provided. This survey provides a comprehensive roadmap to expedite the acclimation process for emerging researchers willing to develop efficient algorithms for detecting ECG anomalies using DL models. Our tailored guidelines bridge the knowledge gap allowing newcomers to align smoothly with the prevailing research trends in ECG arrhythmia detection. We shed light on potential areas for future research and refinement in model development and optimization, intending to stimulate advancement in ECG arrhythmia detection and classification.

A Bibliometric Analysis on Arrhythmia Detection and Classification from 2005 to 2022

Bibliometric analysis is a widely used technique for analyzing large quantities of academic literature and evaluating its impact in a particular academic field. In this paper bibliometric analysis has been used to analyze the academic research on arrhythmia detection and classification from 2005 to 2022. We have followed PRISMA 2020 framework to identify, filter and select the relevant papers. This study has used the Web of Science database to find related publications on arrhythmia detection and classification. "Arrhythmia detection", "arrhythmia classification" and "arrhythmia detection and classification" are three keywords for gathering the relevant articles. 238 publications in total were selected for this research. In this study, two different bibliometric techniques, "performance analysis" and "science mapping", were applied. Different bibliometric parameters such as publication analysis, trend analysis, citation analysis, and networking analysis have been used to evaluate the performance of these articles. According to this analysis, the three countries with the highest number of publications and citations are China, the USA, and India in terms of arrhythmia detection and classification. The three most significant researchers in this field are those named U. R. Acharya, S. Dogan, and P. Plawiak. Machine learning, ECG, and deep learning are the three most frequently used keywords. A further finding of the study indicates that the popular topics for arrhythmia identification are machine learning, ECG, and atrial fibrillation. This research provides insight into the origins, current status, and future direction of arrhythmia detection research.

Detection of Paroxysmal Atrial Fibrillation from Dynamic ECG Recordings Based on a Deep Learning Model

BACKGROUND AND OBJECTIVES

Atrial fibrillation (AF) is one of the most common arrhythmias clinically. Aging tends to increase the risk of AF, which also increases the burden of other comorbidities, including coronary artery disease (CAD), and even heart failure (HF). The precise detection of AF is a challenge due to its intermittence and unpredictability. A method for the accurate detection of AF is still needed.

METHODS

A deep learning model was used to detect atrial fibrillation. Here, a distinction was not made between AF and atrial flutter (AFL), both of which manifest as a similar pattern on an electrocardiogram (ECG). This method not only discriminated AF from normal rhythm of the heart, but also detected its onset and offset. The proposed model involved residual blocks and a Transformer encoder.

RESULTS AND CONCLUSIONS

The data used for training were obtained from the CPSC2021 Challenge, and were collected using dynamic ECG devices. Tests on four public datasets validated the availability of the proposed method. The best performance for AF rhythm testing attained an accuracy of 98.67%, a sensitivity of 87.69%, and a specificity of 98.56%. In onset and offset detection, it obtained a sensitivity of 95.90% and 87.70%, respectively. The algorithm with a low FPR of 0.46% was able to reduce troubling false alarms. The model had a great capability to discriminate AF from normal rhythm and to detect its onset and offset. Noise stress tests were conducted after mixing three types of noise. We visualized the model's features using a heatmap and illustrated its interpretability. The model focused directly on the crucial ECG waveform where showed obvious characteristics of AF.

Automatic screening of patients with atrial fibrillation from 24-h Holter recording using deep learning

Aims

As the demand for atrial fibrillation (AF) screening increases, clinicians spend a significant amount of time identifying AF signals from massive amounts of data obtained during long-term dynamic electrocardiogram (ECG) monitoring. The identification of AF signals is subjective and depends on the experience of clinicians. However, experienced cardiologists are scarce. This study aimed to apply a deep learning-based algorithm to fully automate primary screening of patients with AF using 24-h Holter monitoring.

Methods and results

A deep learning model was developed to automatically detect AF episodes using RR intervals and was trained and evaluated on 23 621 (2297 AF and 21 324 non-AF) 24-h Holter recordings from 23 452 patients. Based on the AF episode detection results, patients with AF were automatically identified using the criterion of at least one AF episode lasting 6 min or longer. Performance was assessed on an independent real-world hospital-scenario test set (19 227 recordings) and a community-scenario test set (1299 recordings). For the two test sets, the model obtained high performance for the identification of patients with AF (sensitivity: 0.995 and 1.000; specificity: 0.985 and 0.997, respectively). Moreover, it obtained good and consistent performance (sensitivity: 1.000; specificity: 0.972) for an external public data set.

Conclusion

Using the criterion of at least one AF episode of 6 min or longer, the deep learning model can fully automatically screen patients for AF with high accuracy from long-term Holter monitoring data. This method may serve as a powerful and cost-effective tool for primary screening for AF.

Speech Emotion Recognition Using Attention Model

Speech emotion recognition is an important research topic that can help to maintain and improve public health and contribute towards the ongoing progress of healthcare technology. There have been several advancements in the field of speech emotion recognition systems including the use of deep learning models and new acoustic and temporal features. This paper proposes a self-attention-based deep learning model that was created by combining a two-dimensional Convolutional Neural Network (CNN) and a long short-term memory (LSTM) network. This research builds on the existing literature to identify the best-performing features for this task with extensive experiments on different combinations of spectral and rhythmic information. Mel Frequency Cepstral Coefficients (MFCCs) emerged as the best performing features for this task. The experiments were performed on a customised dataset that was developed as a combination of RAVDESS, SAVEE, and TESS datasets. Eight states of emotions (happy, sad, angry, surprise, disgust, calm, fearful, and neutral) were detected. The proposed attention-based deep learning model achieved an average test accuracy rate of 90%, which is a substantial improvement over established models. Hence, this emotion detection model has the potential to improve automated mental health monitoring.

Bimodal CNN for cardiovascular disease classification by co-training ECG grayscale images and scalograms

This study aimed to develop a bimodal convolutional neural network (CNN) by co-training grayscale images and scalograms of ECG for cardiovascular disease classification. The bimodal CNN model was developed using a 12-lead ECG database collected from Chapman University and Shaoxing People's Hospital. The preprocessed database contains 10,588 ECG data and 11 heart rhythms labeled by a specialist physician. The preprocessed one-dimensional ECG signals were converted into two-dimensional grayscale images and scalograms, which are fed simultaneously to the bimodal CNN model as dual input images. The proposed model aims to improve the performance of CVDs classification by making use of ECG grayscale images and scalograms. The bimodal CNN model consists of two identical Inception-v3 backbone models, which were pre-trained on the ImageNet database. The proposed model was fine-tuned with 6780 dual-input images, validated with 1694 dual-input images, and tested on 2114 dual-input images. The bimodal CNN model using two identical Inception-v3 backbones achieved best AUC (0.992), accuracy (95.08%), sensitivity (0.942), precision (0.946) and F1-score (0.944) in lead II. Ensemble model of all leads obtained AUC (0.994), accuracy (95.74%), sensitivity (0.950), precision (0.953), and F1-score (0.952). The bimodal CNN model showed better diagnostic performance than logistic regression, XGBoost, LSTM, single CNN model training with grayscale images alone or with scalograms alone. The proposed bimodal CNN model would be of great help in diagnosing cardiovascular diseases.

Multi-Modal Stacking Ensemble for the Diagnosis of Cardiovascular Diseases

BACKGROUND

Cardiovascular diseases (CVDs) are a leading cause of death worldwide. Deep learning methods have been widely used in the field of medical image analysis and have shown promising results in the diagnosis of CVDs.

METHODS

Experiments were performed on 12-lead electrocardiogram (ECG) databases collected by Chapman University and Shaoxing People's Hospital. The ECG signal of each lead was converted into a scalogram image and an ECG grayscale image and used to fine-tune the pretrained ResNet-50 model of each lead. The ResNet-50 model was used as a base learner for the stacking ensemble method. Logistic regression, support vector machine, random forest, and XGBoost were used as a meta learner by combining the predictions of the base learner. The study introduced a method called multi-modal stacking ensemble, which involves training a meta learner through a stacking ensemble that combines predictions from two modalities: scalogram images and ECG grayscale images.

RESULTS

The multi-modal stacking ensemble with a combination of ResNet-50 and logistic regression achieved an AUC of 0.995, an accuracy of 93.97%, a sensitivity of 0.940, a precision of 0.937, and an F1-score of 0.936, which are higher than those of LSTM, BiLSTM, individual base learners, simple averaging ensemble, and single-modal stacking ensemble methods.

CONCLUSION

The proposed multi-modal stacking ensemble approach showed effectiveness for diagnosing CVDs.

Method for Solving Difficulties in Rhythm Classification Caused by Few Samples and Similar Characteristics in Electrocardiograms

A method for accurately analyzing electrocardiograms (ECGs), which are obtained from electrical signals generated by cardiac activity, is essential in heart disease diagnosis. However, rhythms are typically obtained with relatively few data samples and similar characteristics, making them difficult to classify. To solve these issues, we proposed a novel method that distinguishes a given ECG rhythm using a beat score map (BSM) image. Through the proposed method, the associations between beats and previously used features, such as the R-R interval, were considered. Rhythm classification was implemented by training a convolutional neural network model and using transfer learning with the created BSM image. As a result, the proposed method for ECG rhythms with small data samples showed significant results. It also showed good performance in differentiating atrial fibrillation (AFIB) and atrial flutter (AFL) rhythms, which are difficult to distinguish due to their similar characteristics. The performance for rhythms with a small number of samples of the proposed method is 20% better than an existing method. In addition, the performance based on the F-1 score for classifying AFIB and AFL of the proposed method is 30% better than the existing method. This study solved the previous limitations caused by small sample numbers and similar rhythms.

Comparison of ANN and ANFIS Models for AF Diagnosis Using RR Irregularities

Classification of normal sinus rhythm (NSR), paroxysmal atrial fibrillation (PAF), and persistent atrial fibrillation (AF) is crucial in order to diagnose and effectively plan treatment for patients. Current classification models were primarily developed by electrocardiogram (ECG) signal databases, which may be unsuitable for local patients. Therefore, this research collected ECG signals from 60 local Thai patients (age 52.53 ± 23.92) to create a classification model. The coefficient of variance (CV), the median absolute deviation (MAD), and the root mean square of the successive differences (RMSSD) are ordinary feature variables of RR irregularities used by existing models. The square of average variation (SAV) is a newly proposed feature that extracts from the irregularity of RR intervals. All variables were found to be statistically different using ANOVA tests and Tukey’s method with a p-value less than 0.05. The methods of artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) were also tested and compared to find the best classification model. Finally, SAV showed the best performance using the ANFIS model with trapezoidal membership function, having the highest system accuracy (ACC) at 89.33%, sensitivity (SE), specificity (SP), and positive predictivity (PPR) for NSR at 100.00%, 94.00%, and 89.29%, PAF at 88.00%, 90.57%, and 81.48%, and AF at 80.00%, 96.00%, and 90.91%, respectively.

DAE-ConvBiLSTM: End-to-end learning single-lead electrocardiogram signal for heart abnormalities detection

BACKGROUND

The electrocardiogram (ECG) is a widely used diagnostic that observes the heart activities of patients to ascertain a heart abnormality diagnosis. The artifacts or noises are primarily associated with the problem of ECG signal processing. Conventional denoising techniques have been proposed in previous literature; however, some lacks, such as the determination of suitable wavelet basis function and threshold, can be a time-consuming process. This paper presents end-to-end learning using a denoising auto-encoder (DAE) for denoising algorithms and convolutional-bidirectional long short-term memory (ConvBiLSTM) for ECG delineation to classify ECG waveforms in terms of the PQRST-wave and isoelectric lines. The denoising reconstruction using unsupervised learning based on the encoder-decoder process can be proposed to improve the drawbacks. First, The ECG signals are reduced to a low-dimensional vector in the encoder. Second, the decoder reconstructed the signals. The last, the reconstructed signals of ECG can be processed to ConvBiLSTM. The proposed architecture of DAE-ConvBiLSTM is the end-to-end diagnosis of heart abnormality detection.

RESULTS

As a result, the performance of DAE-ConvBiLSTM has obtained an average of above 98.59% accuracy, sensitivity, specificity, precision, and F1 score from the existing studies. The DAE-ConvBiLSTM has also experimented with detecting T-wave (due to ventricular repolarisation) morphology abnormalities.

CONCLUSION

The development architecture for detecting heart abnormalities using an unsupervised learning DAE and supervised learning ConvBiLSTM can be proposed for an end-to-end learning algorithm. In the future, the precise accuracy of the ECG main waveform will affect heart abnormalities detection in clinical practice.

A 3D-CNN with temporal-attention block to predict the recurrence of atrial fibrillation based on body-surface potential mapping signals

Catheter ablation has become an important treatment for atrial fibrillation (AF), but its recurrence rate is still high. The aim of this study was to predict AF recurrence using a three-dimensional (3D) network model based on body-surface potential mapping signals (BSPMs). BSPMs were recorded with a 128-lead vest in 14 persistent AF patients before undergoing catheter ablation (Maze-IV). The torso geometry was acquired and meshed by point cloud technology, and the BSPM was interpolated into the torso geometry by the inverse distance weighted (IDW) method to generate the isopotential map. Experiments show that the isopotential map of BSPMs can reflect the propagation of the electrical wavefronts. The 3D isopotential sequence map was established by combining the spatial–temporal information of the isopotential map; a 3D convolutional neural network (3D-CNN) model with temporal attention was established to predict AF recurrence. Our study proposes a novel attention block that focuses the characteristics of atrial activations to improve sampling accuracy. In our experiment, accuracy (ACC) in the intra-patient evaluation for predicting the recurrence of AF was 99.38%. In the inter-patient evaluation, ACC of 3D-CNN was 81.48%, and the area under the curve (AUC) was 0.88. It can be concluded that the dynamic rendering of multiple isopotential maps can not only comprehensively display the conduction of cardiac electrical activity on the body surface but also successfully predict the recurrence of AF after CA by using 3D isopotential sequence maps.

Atrial fibrillation prediction by combining ECG markers and CMR radiomics

Atrial fibrillation (AF) is the most common cardiac arrhythmia. It is associated with a higher risk of important adverse health outcomes such as stroke and death. AF is linked to distinct electro-anatomic alterations. The main tool for AF diagnosis is the Electrocardiogram (ECG). However, an ECG recorded at a single time point may not detect individuals with paroxysmal AF. In this study, we developed machine learning models for discrimination of prevalent AF using a combination of image-derived radiomics phenotypes and ECG features. Thus, we characterize the phenotypes of prevalent AF in terms of ECG and imaging alterations. Moreover, we explore sex-differential remodelling by building sex-specific models. Our integrative model including radiomics and ECG together resulted in a better performance than ECG alone, particularly in women. ECG had a lower performance in women than men (AUC: 0.77 vs 0.88, p < 0.05) but adding radiomics features, the accuracy of the model was able to improve significantly. The sensitivity also increased considerably in women by adding the radiomics (0.68 vs 0.79, p < 0.05) having a higher detection of AF events. Our findings provide novel insights into AF-related electro-anatomic remodelling and its variations by sex. The integrative radiomics-ECG model also presents a potential novel approach for earlier detection of AF.

A novel deep learning package for electrocardiography research

Objective. In recent years, deep learning has blossomed in the field of electrocardiography (ECG) processing, outperforming traditional signal processing methods in a number of typical tasks; for example, classification, QRS detection and wave delineation. Although many neural architectures have been proposed in the literature, there is a lack of systematic studies and open-source libraries for ECG deep learning.Approach. In this paper, we propose a deep learning package, namedtorch_ecg, which assembles a large number of neural networks, from existing and novel literature, for various ECG processing tasks. The models are designed to be able to be automatically built from configuration files that contain a large set of configurable hyperparameters, making it convenient to scale the networks and perform neural architecture searching.torch_ecghas well-organized data processing modules, which contain utilities for data downloading, visualization, preprocessing and augmentation. To make the whole system more user-friendly, a series of helper modules are implemented, including model trainers, metric computation and loggers.Main results.torch_ecgestablishes a convenient and modular way for automatic building and flexible scaling of networks, as well as a neat and uniform way of organizing the preprocessing procedures and augmentation techniques for preparing the input data for the models. In addition,torch_ecgprovides benchmark studies using the latest databases, illustrating the principles and pipelines for solving ECG processing tasks and reproducing results from the literature.Significance.torch_ecgoffers the ECG research community a powerful tool for meeting the growing demand for the application of deep learning techniques. The code is available athttps://github.com/DeepPSP/torch_ecg.

A comparative study on neural networks for paroxysmal atrial fibrillation events detection from electrocardiography

OBJECTIVE

This work conducts a comparative study on the effect of neural networks of different architectures on the detection of paroxysmal atrial fibrillation (PAF) events from dynamic electrocardiography (ECG) recordings, a problem raised in the 4th China Physiological Signal Challenge 2021 (CPSC2021).

APPROACH

We proposed 3 neural network models and an auxiliary one for QRS detection to tackle the problem. A convolutional recurrent neural network (CRNN) model and a U-Net model that accepts ECG waveform input make sample-wise predictions. This regards the PAF events detection as a segmentation task. A stacked bidirectional long short-term memory (LSTM) model takes the sequence of RR intervals, which is derived from the output of the QRS detection model and makes beat-wise predictions. The QRS detection model also has a CRNN architecture, which is slightly different from the model for the AF segmentation task. Final predictions are merged by outputs from models making sample-wise predictions and making beat-wise predictions. Finally, the locations of QRS complexes are used to filter out segments (both normal and AF) shorter than 5 beats. In order to make the neural network models more sensitive to the critical sample points (onsets and offsets) of the AF events, we proposed a novel masked binary cross-entropy (MaskedBCE) loss function for training the models. This loss function is the conventional BCE loss multiplied by a mask, whose values in a neighbourhood of critical sample points are significantly larger than elsewhere.

MAIN RESULTS

Our method received a score of 1.9972 on the first part of the hidden test set of CPSC2021 and a score of 3.0907 on the second part. The average score was 2.5440, ranked 5th out of 17 teams with successful official entries.

SIGNIFICANCE

This work proposed an effective solution to the problem of the detection of PAF events from dynamic ECGs and validated the efficacy of several neural network architectures on this problem.

An automated detection of atrial fibrillation from single‑lead ECG using HRV features and machine learning

BACKGROUND

Atrial fibrillation (AF) is a disorder of the heart rhythm where irregular and rapid heartbeats are observed. This supraventricular arrhythmia may increase the risk of blood clots, stroke, heart failure, and other serious heart complications. Automatic analysis of AF that is based on machine learning (ML) plays an important role in detecting this heart disease.

METHODS

A new approach for automated AF detection is presented using heart rate variability (HRV) features and machine learning. A set of time-domain, frequency-domain and nonlinear features are extracted from the R-R intervals. A new method for frequency-domain analysis of R-R intervals using the Fourier Decomposition Method is presented, which provides promising results as compared to the usual method of power spectral density estimation. We train the algorithm on the Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) atrial fibrillation database and perform a comprehensive analysis using statistical tests to obtain the results without any intra-patient bias.

RESULTS

The proposed method is able to achieve average result of 95.16% sensitivity, 92.46% specificity and 94.43% accuracy and its performance is better than the existing approaches. Furthermore, the efficacy of the proposed algorithm is tested on eight records from a previously unseen MIT-BIH Arrhythmia Database.

CONCLUSION

This work shows that the proposed HRV features and ML approach can be effectively used for the analysis, detection, and classification of AF.

Automated diagnosis of atrial fibrillation using ECG component-aware transformer

Atrial fibrillation (AF) is the most common sustained arrhythmia worldwide and imposes a substantial economic burden on the public healthcare system due to its high morbidity and mortality. Early detection of AF is crucial in providing timely treatment and preventing complications such as stroke and other thromboembolism. For AF diagnosis, the 12-lead electrocardiogram (ECG) has been established as the gold standard. However, it requires the clinical experiences of cardiologists and may be vulnerable to inter-observer variability. Although automated AF diagnostic techniques based on deep neural networks (DNN) have been proposed, most studies were conducted using small-scale datasets, resulting in the over-fitting problem. Furthermore, they have not fully exploited ECG components such as P-wave, QRS-complex, and T-wave contrary to the approach adopted by cardiologists who interpret ECG by considering its components. To overcome these limitations, this study presents the component-aware transformer (CAT), which segments the ECG waveform into each component, vectorizes them with length and types information into one vector, and used it as the input of the transformer. We conducted extensive experiments to evaluate the CAT using a large-scale dataset called Shaoxing Hospital Zhejiang University School of Medicine database (AF: 1,780 cases, non-AF: 8,866 cases). The quantitative evaluations demonstrate that the CAT outperforms the conventional deep learning techniques on both single- and 12-lead ECG signals. Moreover, the CAT trained on single-lead ECG is comparable to that of a 12-lead analysis, while conventional methods degraded significantly in performance. Consequently, the CAT is applicable to various single-channel signals such as airway pressure, photoplethysmogram, and blood pressure.

Applying Artificial Intelligence to Wearable Sensor Data to Diagnose and Predict Cardiovascular Disease: A Review

Cardiovascular disease (CVD) is the world's leading cause of mortality. There is significant interest in using Artificial Intelligence (AI) to analyse data from novel sensors such as wearables to provide an earlier and more accurate prediction and diagnosis of heart disease. Digital health technologies that fuse AI and sensing devices may help disease prevention and reduce the substantial morbidity and mortality caused by CVD worldwide. In this review, we identify and describe recent developments in the application of digital health for CVD, focusing on AI approaches for CVD detection, diagnosis, and prediction through AI models driven by data collected from wearables. We summarise the literature on the use of wearables and AI in cardiovascular disease diagnosis, followed by a detailed description of the dominant AI approaches applied for modelling and prediction using data acquired from sensors such as wearables. We discuss the AI algorithms and models and clinical applications and find that AI and machine-learning-based approaches are superior to traditional or conventional statistical methods for predicting cardiovascular events. However, further studies evaluating the applicability of such algorithms in the real world are needed. In addition, improvements in wearable device data accuracy and better management of their application are required. Lastly, we discuss the challenges that the introduction of such technologies into routine healthcare may face.

Multiscale Encoding of Electrocardiogram Signals with a Residual Network for the Detection of Atrial Fibrillation

Atrial fibrillation (AF) is one of the most common cardiac arrhythmias, and it is an indication of high-risk factors for stroke, myocardial ischemia, and other malignant cardiovascular diseases. Most of the existing AF detection methods typically convert one-dimensional time-series electrocardiogram (ECG) signals into two-dimensional representations to train a deep and complex AF detection system, which results in heavy training computation and high implementation costs. In this paper, a multiscale signal encoding scheme is proposed to improve feature representation and detection performance without the need for using any transformation or handcrafted feature engineering techniques. The proposed scheme uses different kernel sizes to produce the encoded signal by using multiple streams that are passed into a one-dimensional sequence of blocks of a residual convolutional neural network (ResNet) to extract representative features from the input ECG signal. This also allows networks to grow in breadth rather than in depth, thus reducing the computing time by using the parallel processing capability of deep learning networks. We investigated the effects of the use of a different number of streams with different kernel sizes on the performance. Experiments were carried out for a performance evaluation using the publicly available PhysioNet CinC Challenge 2017 dataset. The proposed multiscale encoding scheme outperformed existing deep learning-based methods with an average F₁ score of 98.54%, but with a lower network complexity.

Quantifying the impact of Pyramid Squeeze Attention mechanism and filtering approaches on Alzheimer's disease classification

Brain medical imaging and deep learning are important foundations for diagnosing and predicting Alzheimer's disease. In this study, we explored the impact of different image filtering approaches and Pyramid Squeeze Attention (PSA) mechanism on the image classification of Alzheimer's disease. First, during the image preprocessing, we register MRI images and remove skulls, then apply median filtering, Gaussian blur filtering, and anisotropic diffusion filtering to obtain different experimental images. After that, we add the Squeeze and Excitation (SE) mechanism and Pyramid Squeeze Attention (PSA) mechanism to the Fully Convolutional Network (FCN) model respectively, to obtain each MRI image's corresponding feature information of disease probability map. Besides, we also construct Multi-Layer Perceptron (MLP) model's framework, combining feature information of disease probability map with age, gender, and Mini-Mental State Examination (MMSE) of each sample, to get the final classification performance of model. Among them, the accuracy of the MLP-C model combining anisotropic diffusion filtering with the Pyramid Squeeze Attention mechanism can reach 98.85%. The corresponding quantitative experimental results show that different image filtering approaches and attention mechanisms provide effective assistance for the diagnosis and classification of Alzheimer's disease.

Atrioventricular Synchronization for Detection of Atrial Fibrillation and Flutter in One to Twelve ECG Leads Using a Dense Neural Network Classifier

This study investigates the use of atrioventricular (AV) synchronization as an important diagnostic criterion for atrial fibrillation and flutter (AF) using one to twelve ECG leads. Heart rate, lead-specific AV conduction time, and P-/f-wave amplitude were evaluated by three representative ECG metrics (mean value, standard deviation), namely RR-interval (RRi-mean, RRi-std), PQ-interval (PQi-mean, PQI-std), and PQ-amplitude (PQa-mean, PQa-std), in 71,545 standard 12-lead ECG records from the six largest PhysioNet CinC Challenge 2021 databases. Two rhythm classes were considered (AF, non-AF), randomly assigning records into training (70%), validation (20%), and test (10%) datasets. In a grid search of 19, 55, and 83 dense neural network (DenseNet) architectures and five independent training runs, we optimized models for one-lead, six-lead (chest or limb), and twelve-lead input features. Lead-set performance and SHapley Additive exPlanations (SHAP) input feature importance were evaluated on the test set. Optimal DenseNet architectures with the number of neurons in sequential [1st, 2nd, 3rd] hidden layers were assessed for sensitivity and specificity: DenseNet [16,16,0] with primary leads (I or II) had 87.9-88.3 and 90.5-91.5%; DenseNet [32,32,32] with six limb leads had 90.7 and 94.2%; DenseNet [32,32,4] with six chest leads had 92.1 and 93.2%; and DenseNet [128,8,8] with all 12 leads had 91.8 and 95.8%, indicating sensitivity and specificity values, respectively. Mean SHAP values on the entire test set highlighted the importance of RRi-mean (100%), RR-std (84%), and atrial synchronization (40-60%) for the PQa-mean (aVR, I), PQi-std (V2, aVF, II), and PQi-mean (aVL, aVR). Our focus on finding the strongest AV synchronization predictors of AF in 12-lead ECGs would lead to a comprehensive understanding of the decision-making process in advanced neural network classifiers. DenseNet self-learned to rely on a few ECG behavioral characteristics: first, characteristics usually associated with AF conduction such as rapid heart rate, enhanced heart rate variability, and large PQ-interval deviation in V2 and inferior leads (aVF, II); second, characteristics related to a typical P-wave pattern in sinus rhythm, which is best distinguished from AF by the earliest negative P-peak deflection of the right atrium in the lead (aVR) and late positive left atrial deflection in lateral leads (I, aVL). Our results on lead-selection and feature-selection practices for AF detection should be considered for one- to twelve-lead ECG signal processing settings, particularly those measuring heart rate, AV conduction times, and P-/f-wave amplitudes. Performances are limited to the AF diagnostic potential of these three metrics. SHAP value importance can be used in combination with a human expert's ECG interpretation to change the focus from a broad observation of 12-lead ECG morphology to focusing on the few AV synchronization findings strongly predictive of AF or non-AF arrhythmias. Our results are representative of AV synchronization findings across a broad taxonomy of cardiac arrhythmias in large 12-lead ECG databases.

Accurate detection of atrial fibrillation events with R-R intervals from ECG signals

Atrial fibrillation (AF) is a typical category of arrhythmia. Clinical diagnosis of AF is based on the detection of abnormal R-R intervals (RRIs) with an electrocardiogram (ECG). Previous studies considered this detection problem as a classification problem and focused on extracting a number of features. In this study we demonstrate that instead of using any specific numerical characteristic as the input feature, the probability density of RRIs from ECG conserves comprehensive statistical information; hence, is a natural and efficient input feature for AF detection. Incorporated with a support vector machine as the classifier, results on the MIT-BIH database indicates that the proposed method is a simple and accurate approach for AF detection in terms of accuracy, sensitivity, and specificity.

MGNN: A multiscale grouped convolutional neural network for efficient atrial fibrillation detection

The reliable detection of atrial fibrillation (AF) is of great significance for monitoring disease progression and developing tailored care paths. In this work, we proposed a novel and robust method based on deep learning for the accurate detection of AF. Using RR interval sequences, a multiscale grouped convolutional neural network (MGNN) combined with self-attention was designed for automatic feature extraction, and AF and non-AF classification. An average accuracy of 97.07% was obtained in the 5-fold cross-validation. The generalization ability of the proposed MGNN was further independently tested on four other unseen datasets, and the accuracy was 92.23%, 96.86%, 94.23% and 95.91%. Moreover, comparison of the network structures indicated that the MGNN had not only better detection performance but also lower computational complexity. In conclusion, the proposed model is shown to be an efficient AF detector that has great potential for use in clinical auxiliary diagnosis and long-term home monitoring based on wearable devices.

Using Minimum Redundancy Maximum Relevance Algorithm to Select Minimal Sets of Heart Rate Variability Parameters for Atrial Fibrillation Detection

Heart rate is quite regular during sinus (normal) rhythm (SR) originating from the sinus node. In contrast, heart rate is usually irregular during atrial fibrillation (AF). Complete atrioventricular block with an escape rhythm, ventricular pacing, or ventricular tachycardia are the most common exceptions when heart rate may be regular in AF. Heart rate variability (HRV) is the variation in the duration of consecutive cardiac cycles (RR intervals). We investigated the utility of HRV parameters for automated detection of AF with machine learning (ML) classifiers. The minimum redundancy maximum relevance (MRMR) algorithm, one of the most effective algorithms for feature selection, helped select the HRV parameters (including five original), best suited for distinguishing AF from SR in a database of over 53,000 60 s separate electrocardiogram (ECG) segments cut from longer (up to 24 h) ECG recordings. HRV parameters entered the ML-based classifiers as features. Seven different, commonly used classifiers were trained with one to six HRV-based features with the highest scores resulting from the MRMR algorithm and tested using the 5-fold cross-validation and blindfold validation. The best ML classifier in the blindfold validation achieved an accuracy of 97.2% and diagnostic odds ratio of 1566. From all studied HRV features, the top three HRV parameters distinguishing AF from SR were: the percentage of successive RR intervals differing by at least 50 ms (pRR50), the ratio of standard deviations of points along and across the identity line of the Poincare plots, respectively (SD2/SD1), and coefficient of variation—standard deviation of RR intervals divided by their mean duration (CV). The proposed methodology and the presented results of the selection of HRV parameters have the potential to develop practical solutions and devices for automatic AF detection with minimal sets of simple HRV parameters. Using straightforward ML classifiers and the extremely small sets of simple HRV features, always with pRR50 included, the differentiation of AF from sinus rhythms in the 60 s ECGs is very effective.

A systematic review and Meta-data analysis on the applications of Deep Learning in Electrocardiogram

The success of deep learning over the traditional machine learning techniques in handling artificial intelligence application tasks such as image processing, computer vision, object detection, speech recognition, medical imaging and so on, has made deep learning the buzz word that dominates Artificial Intelligence applications. From the last decade, the applications of deep learning in physiological signals such as electrocardiogram (ECG) have attracted a good number of research. However, previous surveys have not been able to provide a systematic comprehensive review including biometric ECG based systems of the applications of deep learning in ECG with respect to domain of applications. To address this gap, we conducted a systematic literature review on the applications of deep learning in ECG including biometric ECG based systems. The study analyzed systematically, 150 primary studies with evidence of the application of deep learning in ECG. The study shows that the applications of deep learning in ECG have been applied in different domains. We presented a new taxonomy of the domains of application of the deep learning in ECG. The paper also presented discussions on biometric ECG based systems and meta-data analysis of the studies based on the domain, area, task, deep learning models, dataset sources and preprocessing methods. Challenges and potential research opportunities were highlighted to enable novel research. We believe that this study will be useful to both new researchers and expert researchers who are seeking to add knowledge to the already existing body of knowledge in ECG signal processing using deep learning algorithm.

Supplementary information

The online version contains supplementary material available at 10.1007/s12652-022-03868-z.

Automatic Atrial Fibrillation Arrhythmia Detection Using Univariate and Multivariate Data

Atrial fibrillation (AF) is still a major cause of disease morbidity and mortality, making its early diagnosis desirable and urging researchers to develop efficient methods devoted to automatic AF detection. Till now, the analysis of Holter-ECG recordings remains the gold-standard technique to screen AF. This is usually achieved by studying either RR interval time series analysis, P-wave detection or combinations of both morphological characteristics. After extraction and selection of meaningful features, each of the AF detection methods might be conducted through univariate and multivariate data analysis. Many of these automatic techniques have been proposed over the last years. This work presents an overview of research studies of AF detection based on RR interval time series. The aim of this paper is to provide the scientific community and newcomers to the field of AF screening with a resource that presents introductory concepts, clinical features, and a literature review that describes the techniques that are mostly followed when RR interval time series are used for accurate detection of AF.

Investigation of the Evolution of Wavelet Higher-Order Dynamics in Atrial Fibrillation

Atrial fibrillation (AF) is the most common cardiac arrhythmia and is associated with significant morbidity and mortality. Owing to the advances in sensor technology and the emergence of wearable devices that enable daily self-monitoring, ECG signal processing methods for the automatic detection of AF are more pertinent than ever. In this paper, we investigate the use of wavelet higher-order statistics (WHOS) for feature extraction and differentiation between normal sinus rhythm and AF. The proposed approach captures the evolution of the WHOS dynamics and quantifies the changes in the time-varying characteristics of the frequency couplings caused by AF. Results obtained from the statistical analysis of a dataset of 5834 single-lead ECG recordings, reveal 46/50 statistically significant features and provide insight into the complexity of the evolution of the ECG non-linearities during AF.

ECG Classification Using Orthogonal Matching Pursuit and Machine Learning

Health monitoring and related technologies are a rapidly growing area of research. To date, the electrocardiogram (ECG) remains a popular measurement tool in the evaluation and diagnosis of heart disease. The number of solutions involving ECG signal monitoring systems is growing exponentially in the literature. In this article, underestimated Orthogonal Matching Pursuit (OMP) algorithms are used, demonstrating the significant effect of concise representation parameters on improving the performance of the classification process. Cardiovascular disease classification models based on classical Machine Learning classifiers were defined and investigated. The study was undertaken on the recently published PTB-XL database, whose ECG signals were previously subjected to detailed analysis. The classification was realized for class 2, class 5, and class 15 cardiac diseases. A new method of detecting R-waves and, based on them, determining the location of QRS complexes was presented. Novel aggregation methods of ECG signal fragments containing QRS segments, necessary for tests for classical classifiers, were developed. As a result, it was proved that ECG signal subjected to algorithms of R wave detection, QRS complexes extraction, and resampling performs very well in classification using Decision Trees. The reason can be found in structuring the signal due to the actions mentioned above. The implementation of classification issues achieved the highest Accuracy of 90.4% in recognition of 2 classes, as compared to less than 78% for 5 classes and 71% for 15 classes.

A Micro Neural Network for Healthcare Sensor Data Stream Classification in Sustainable and Smart Cities

A smart city is an intelligent space, in which large amounts of data are collected and analyzed using low-cost sensors and automatic algorithms. The application of artificial intelligence and Internet of Things (IoT) technologies in electronic health (E-health) can efficiently promote the development of sustainable and smart cities. The IoT sensors and intelligent algorithms enable the remote monitoring and analyzing of the healthcare data of patients, which reduces the medical and travel expenses in cities. Existing deep learning-based methods for healthcare sensor data classification have made great achievements. However, these methods take much time and storage space for model training and inference. They are difficult to be deployed in small devices to classify the physiological signal of patients in real time. To solve the above problems, this paper proposes a micro time series classification model called the micro neural network (MicroNN). The proposed model is micro enough to be deployed on tiny edge devices. MicroNN can be applied to long-term physiological signal monitoring based on edge computing devices. We conduct comprehensive experiments to evaluate the classification accuracy and computation complexity of MicroNN. Experiment results show that MicroNN performs better than the state-of-the-art methods. The accuracies on the two datasets (MIT-BIH-AR and INCART) are 98.4% and 98.1%, respectively. Finally, we present an application to show how MicroNN can improve the development of sustainable and smart cities.

Deepaware: A hybrid deep learning and context-aware heuristics-based model for atrial fibrillation detection

BACKGROUND

State-of-the-art automatic atrial fibrillation (AF) detection models trained on RR-interval (RRI) features generally produce high performance on standard benchmark electrocardiogram (ECG) AF datasets. These models, however, result in a significantly high false positive rates (FPRs) when applied on ECG data collected under free-living ambulatory conditions and in the presence of non-AF arrhythmias.

METHOD

This paper proposes DeepAware, a novel hybrid model combining deep learning (DL) and context-aware heuristics (CAH), which reduces the FPR effectively and improves the AF detection performance on participant-operated ambulatory ECG from free-living conditions. It exploits the RRI and P-wave features, as well as the contextual features from the ambulatory ECG.

RESULTS

DeepAware is shown to be very generalizable and superior to the state-of-the-art models when applied on unseen benchmark ECG AF datasets. Most importantly, the model is able to detect AF efficiently when applied on participant-operated ambulatory ECG recordings from free-living conditions and has achieved a sensitivity (Se), specificity (Sp), and accuracy (Acc) of 97.94%, 98.39%, 98.06%, respectively. Results also demonstrate the effect of atrial activity analysis (via P-waves detection) and CAH in reducing the FPR over the RRI features-based AF detection model.

CONCLUSIONS

The proposed DeepAware model can substantially reduce the physician's workload of manually reviewing the false positives (FPs) and facilitate long-term ambulatory monitoring for early detection of AF.

Semi-supervised learning for automatic atrial fibrillation detection in 24-hour Holter monitoring

Paroxysmal atrial fibrillation (AF) is generally diagnosed by long-term dynamic electrocardiogram (ECG) monitoring. Identifying AF episodes from long-term ECG data can place a heavy burden on clinicians. Many machine-learning-based automatic AF detection methods have been proposed to solve this issue. However, these methods require numerous annotated data to train the model, and the annotation of AF in long-term ECG is extremely time-consuming. Reducing the demand for labeled data can effectively improve the clinical practicability of automatic AF detection methods. In this study, we developed a novel semi-supervised learning method that generated modified low-entropy labels of unlabeled samples for training a deep learning model to automatically detect paroxysmal AF in 24 h Holter monitoring data. Our method employed a 1D CNN-LSTM neural network with RR intervals as input and used few labeled training data with numerous unlabeled data for training the neural network. This method was evaluated using a 24 h Holter monitoring dataset collected from 1000 paroxysmal AF patients. Using labeled samples from only 10 patients for model training, our method achieved a sensitivity of 97.8%, specificity of 97.9%, and accuracy of 97.9% in five-fold cross-validation. Compared to the supervised learning method with complete labeled samples, the detection accuracy of our method was only 0.5% lower, while the workload of data annotation was significantly reduced by more than 98%. In general, this is the first study to apply semi-supervised learning techniques for automatic AF detection using ECG. Our method can effectively reduce the demand for AF data annotations and can improve the clinical practicability of automatic AF detection.

Effect of temporal resolution on the detection of cardiac arrhythmias using HRV features and machine learning

Objective.Arrhythmia is an abnormal cardiac rhythm that affects the pattern and rate of the heartbeat. Wearable devices with the functionality to measure and store heart rate (HR) data are growing in popularity and enable diagnosing and monitoring arrhythmia on a large scale. The typical sampling resolution of HR data available from non-medical grade wearable devices varies from seconds to several minutes depending on the device and its settings. However, the impact of sampling resolution on the performance and quality of arrhythmia detection has not yet been quantified.Approach.In this study, we investigated the detection and classification of three arrhythmias, namely atrial fibrillation, bradycardia, tachycardia, from down-sampled HR data with various temporal resolution (5-, 15-, 30- and 60 s averages) in 1 h segments extracted from an annotated Holter ECG database acquired at the University of Virginia Heart Station. For the classification task, a total of 15 common heart rate variability (HRV) features were engineered based on the HR time series of each patient. Three different types of machine learning classifiers were evaluated, namely logistic regression, support vector machine and random forest.Main results.A decrease in temporal resolution drastically impacted the detection of atrial fibrillation but did not substantially affect the detection of bradycardia and tachycardia. A HR resolution up to 15 s average demonstrated reasonable performance with a sensitivity of 0.92 and a specificity of 0.86 for a multiclass random forest classifier.Significance.HRV features extracted from low resolution long HR recordings have the potential to increase the early detection of arrhythmias in undiagnosed individuals.

Premature Beats Rejection Strategy on Paroxysmal Atrial Fibrillation Detection

Paroxysmal atrial fibrillation (PAF) may related to the risk of thromboembolism and is the most common cardiac risk factor of cryptogenic stroke (CS). Due to its paroxysmal characteristics, it is usually diagnosed by continuous long-term ECG. Patients with paroxysmal atrial fibrillation usually have premature beats at the same time which is easy to be confused with the rhythm of atrial fibrillation. Therefore, in this article, we designed a screening algorithm for single premature beat, multi premature beats, bigeminy and trigeminy premature beats, according to their rhythm characteristics to reduce false detection caused by premature beats during the PAF detection process. The proposed elimination method was verified on ECG segments with different types of premature beats, and tested on long-term ECG data of PAF patients. ECG segments of different kinds of premature beats were selected from MIT Atrial Fibrillation database (MIT-AFDB), MIT-BIH Arrhythmia database (MIT-AR) and wearable ECG data from the China Physiological Signal Challenge 2021 (CPSC 2021). The proposed method can effectively eliminate single premature beat segments with 99.5% accuracy, and it also can eliminate more than 95% of ECG segments with other types of premature beats. We designed PAF-score as a new index to evaluate the accuracy of detection, and we also calculate the misjudged and missed segments to comprehensively evaluate the PAF detection algorithm. The proposed method get a PAF-score of 0.912 on MIT-AFDB. The proposed method also has the potential to implant low computing power wearable devices for real-time analysis.

An Embedded System Using Convolutional Neural Network Model for Online and Real-Time ECG Signal Classification and Prediction

This paper presents an automatic ECG signal classification system that applied the Deep Learning (DL) model to classify four types of ECG signals. In the first part of our work, we present the model development. Four different classes of ECG signals from the PhysioNet open-source database were selected and used. This preliminary study used a Deep Learning (DL) technique namely Convolutional Neural Network (CNN) to classify and predict the ECG signals from four different classes: normal, sudden death, arrhythmia, and supraventricular arrhythmia. The classification and prediction process includes pulse extraction, image reshaping, training dataset, and testing process. In general, the training accuracy achieved up to 95% after 100 epochs. However, the prediction of each ECG single type shows a differentiation. Among the four classes, the results show that the predictions for sudden death ECG waveforms are the highest, i.e., 80 out of 80 samples are correct (100% accuracy). In contrast, the lowest is the prediction for normal sinus ECG waveforms, i.e., 74 out of 80 samples are correct (92.5% accuracy). This is due to the image features of normal sinus ECG waveforms being almost similar to the image features of supraventricular arrhythmia ECG waveforms. However, the model has been tuned to achieve an optimal prediction. In the second part, we presented the hardware implementation with the predictive model embedded in an NVIDIA Jetson Nanoprocessor for the online and real-time classification of ECG waveforms.

Dashboard Layout Effects on Drivers' Searching Performance and Heart Rate: Experimental Investigation and Prediction

Carsharing scale has been increasing rapidly with sharing economy. However, many users are reluctant to rent cars any longer due to the low-quality of interactive experience and usability, especially in terms of the dashboard design. This challenge should be urgently addressed in order to maintain the sustainable development of car-sharing industry and its environmental benefits. This study aims to investigate the relationship between users' driving activities (e.g., searching time, reading time, eye movement, heart rate) and dashboard layout. This study was conducted based on the experimental investigation among 58 respondents who were required to complete driving tasks in four types of cars with different dashboard layouts. Afterwards, a prediction model was developed to predict users heart rate (HR) based on the long short-term memory model, and logistic models were used to examine the relationship between the occurrence probability of minimum HR and dashboard reading. The results showed that the system usability of a dashboard was related to the drivers' eye movement characteristics including fixation duration, fixation times and pupil diameter. Most indicators had significant effects (p &lt; 0.05) on the system usability score of corresponding dashboard. The long short-term memory model network (RMSE = 1.105, MAE = 0.009) was capable of predicting heart rate (HR) that happened in the process of instrument reading, which presented a periodic pattern rather than a continuous increase or decrease. It reflected that the network could better fit the non-linear and time-sequential laws of HR data. Furthermore, the probability of the lowest heart rate occurrence during the interaction with four dashboards was influenced by the average searching time, reading time and reading accuracy that were related to a specific layout. Overall, this study provided a theoretical reference for uncovering users' adaptive behaviors with the central control screen and for the optimal choice of a suitable dashboard layout in interface design.