Why the study?
Manual ECG signal classification relies heavily on experienced clinicians, consuming substantial time and risking delays in optimal treatment.
Does a 10-layer 1D CNN improve ECG signal classification accuracy compared to existing deep learning methods in the MIT-BIH arrhythmia database?
Population
ECG signals from the MIT-BIH arrhythmia database
Comparison
10-layer one-dimensional convolutional neural network vs other mentioned methods
Key result
A 10-layer 1D convolutional neural network achieved an overall accuracy of 99.43%, sensitivity of 97.86%, and specificity of 99.64% in classifying ECG signals into five categories.
Authors
Loading...
Supports CNN-based ECG tools in research; leaves open clinical use pending prospective validation.
Does a 10-layer 1D CNN improve ECG signal classification accuracy compared to existing deep learning methods in the MIT-BIH arrhythmia database?
A 10-layer 1D CNN model effectively classifies ECG signals into five categories with over 99% accuracy, offering a highly accurate computational tool for automated arrhythmia detection.
Zejun Hu (2023) studied Arrhythmia (n=47). 10-layer 1D Convolutional Neural Network (CNN) vs. Other deep learning methods was evaluated on Classification accuracy. A 10-layer 1D convolutional neural network achieved an overall accuracy of 99.43%, sensitivity of 97.86%, and specificity of 99.64% in classifying ECG signals into five categories.