Exploration and Categorization of Electrocardiogram Signals through the Application of Edge Computing
Keywords:Electrocardiogram (ECG), Signal Analysis, Edge Computing, Machine Learning, Deep Learning
This article presents a comprehensive analysis and categorization of Electrocardiogram (ECG) signals, leveraging the power of edge computing. The study focuses on enhancing the performance of a 12-channel ECG arrhythmia categorization system by implementing advanced oversampling and undersampling techniques. Notably, Random Oversampling (ROS) emerges as particularly effective, demonstrating significant improvements over models trained with original data. Despite commendable performance across various cases, ROS stands out for its straightforward implementation, making it a preferred choice for robust ECG signal categorization. The findings contribute to the refinement of ECG signal processing methodologies, especially in the context of edge computing applications.