Key result
Modified parametric AlexNet outperforms state-of-the-art methods for ECG classification with ~99% accuracy.
Why the study?
Patients with cardiovascular disease require constant monitoring via ECG signal analysis, motivating the development of deep learning networks to extract features and improve disease classification.
Does a modified parametric-based AlexNet structure improve the classification accuracy of ECG signals compared to state-of-the-art methods?
Does a modified parametric-based AlexNet structure improve the classification accuracy of ECG signals compared to state-of-the-art methods?
A modified AlexNet deep learning model using CWT-converted RGB images achieved high accuracy (98.82%) in classifying ECG signals for cardiovascular diseases.
Patients with cardiovascular disease typically need constant monitoring, and this is made possible by analyzing their electrocardiogram (ECG) signals to determine the specific pattern of variation associated with each disease. The various malfunctions of the internal parts of the organ generates multiple ECG signals with different commonalities, which can then be used to classify the sub-categories of each disease. In this study, a deep learning network for classifying ECG signals were presented and features were extracted to improve disease classification. Totally 5,655 ECG recordings was used. Initially all 30-s signals were converted to RGB images using Continuous Wavelet Transform (CWT), and then we those images were fed to AlexNet that was trained using slightly modified parameters. According to the results, the proposed method not only outperforms the state-of-the-art methods (98.82% accuracy) but also has higher recall (98.9%) and precision (97.9%) in the aggregate.
No takes yet. Share an insight, caveat, or question.
Aarthy et al. (2023) studied Cardiovascular diseases (n=5,655). Modified parametric-based AlexNet structure vs. State-of-the-art methods was evaluated on Classification accuracy. A modified parametric-based AlexNet structure for classifying ECG signals achieved 98.82% accuracy, 98.9% recall, and 97.9% precision, outperforming state-of-the-art methods.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: