Identify Failed ECGs and Arrythmia with 1 - D Convolutional Neural Networks

This study addresses an issue with ECGs in which a failed test cannot be recognized until the ECG is printed. Due to a variety of environmental variables and motion artifacts, the ECG signal may get distorted, resulting in erroneous judgments and predictions. As such, getting an uncorrupted and noise - free ECG recording is critical for determining the precise nature of the cardiac problem. This study employs a new technique to enhance the accuracy of an existing method presented by the researcher in another article published a few years ago. A 1 - D Convolutional Neural Network will be used to validate the quality of the image. Following that, the verified ECG will be analyzed to determine the presence of arrhythmia. The results show that this technique increased the accuracy to 99 percent when discriminating between corrupted and normal ECG data and to 93 percent when detecting arrhythmias.