Abstract:Background and objective: Heart disease seriously threatens human life and health. It has the character of abruptness and is necessary to accurately monitor and intelligently diagnose electrocardiograph signals in real-time. As part of the automation of heart monitoring, the electrocardiogram (ECG) intelligent diagnosis method based on deep learning not only meets the needs of real-time and accurate but also can abandon relevant professional knowledge, which makes it possible to be promoted in the general population. Methods: This paper presents an intelligent diagnosis method based on a ResNet. Firstly, ECG signals from MIT-BIH Database are converted into 2-dim matrices by Markov Transition Field. Secondly, the matrices are used as the input of a ResNet. Then, the ResNet is able to extract high abstract features of various diseases and realize intelligent identification of five heartbeat types, including Normal Beat, Left Bundle Branch Block Beat, Right Bundle Branch Block Beat, Premature Ventricular Contraction Beat, and Atrial Premature Contraction Beat. Eventually, the proposed model is used to identify Normal Beat and Atrial Fibrillation(AF) based on the PAF Prediction Challenge Database(the PAFPC Database) to verify its generalization ability. Results: The experiment result shows that the intelligent diagnosis method can reach a high F1-score of 97.7% and a high accuracy upon to 99.2% on MIT-BIH Database, which are higher than the models proposed by other researchers. Its mean sensitivity and mean specificity are 97.42% and 99.54%, respectively. Moreover, the accuracy of the generalization ability verification experiment is 94.57% on the PAFPC Database, which is also higher than the results of other studies. Conclusion: The research results show that the method proposed in this paper still achieves higher accuracy and higher F1-score than other methods without any data preprocessing. This method has better classification performance than traditional machine learning methods and other deep learning methods. That is, the method based on Markov Transition Field and a ResNet has good application prospects. At the same time, it has been verified that the model proposed in this paper also has excellent generalization ability.

Deep learning assists early-detection of hypertension-mediated heart change on ECG signals

Automated Detection of Cardiovascular Disease by Electrocardiogram Signal Analysis: a Deep Learning System

Multi-label Classification of Abnormalities in 12-Lead ECG Using Deep Learning.

Decoding 2.3 million ECGs: interpretable deep learning for advancing cardiovascular diagnosis and mortality risk stratification

A Novel Deep Learning Approach for Early Detection of Cardiovascular Diseases from ECG Signals

An automated ECG-based deep learning for the early-stage identification and classification of cardiovascular disease

A Wearable ECG Monitor for Deep Learning Based Real-Time Cardiovascular Disease Detection

Diagnostic performance of single-lead electrocardiograms for arterial hypertension diagnosis: a machine learning approach

Cardiologist-level interpretable knowledge-fused deep neural network for automatic arrhythmia diagnosis

ECG Signal-Enabled Automatic Diagnosis Technology of Heart Failure

Application of Deep Neural Network for Congestive Heart Failure Detection Using ECG Signals

Detection of Cardiovascular Diseases in ECG Images Using Machine Learning and Deep Learning Methods

An Early Warning of Atrial Fibrillation Based on Short-Time ECG Signals

Using Multi-Task Learning-Based Framework to Detect ST-Segment and J-Point Deviation From Holter

Automatic Diagnosis with 12-Lead ECG Signals.

ECG Arrhythmia Detection Using Disease-specific Attention-based Deep Learning Model

An intelligent diagnostic method of ECG signal based on Markov transition field and a ResNet

Automatic Multi-Label Diagnosis of Single-Lead ECG Using Novel Hybrid Residual Recurrent Convolutional Neural Networks

Automated Detection of Myocardial Infarction and Heart Conduction Disorders Based on Feature Selection and a Deep Learning Model