Decoding Abnormal Heart Patterns: ECG Anomaly Detection Using MobileNet50 CNN Autoencoder Method

In this research investigation, the application of Convolutional Neural Network (MobileN et50) Autoencoders is explored for the detection of anomalies in electrocardiogram (ECG) data using the PTB Diagnostic ECG Database. The dataset comprises 14,552 samples categorized into normal heartbeats and those affected by cardiac abnormalities. The incorporation of Transposed Convolution significantly improves the model's performance. The study underscores the timely recognition of cardiac abnormalities and introduces a specialized MobileNet50 Autoencoder model designed to efficiently encode and decode ECG data, aiding in the identification of irregular patterns. The methodology involves constructing a robust Autoencoder with encoder and decoder components trained to minimize reconstruction errors. Evaluation metrics showcase the model's outstanding accuracy (76.93%), precision (55.23%), recall (89.81%), and F1 score (65.40%). This research underscores the importance of leveraging deep learning techniques for early anomaly detection in ECG data, presenting a promising avenue for enhancing diagnostic capabilities and ultimately benefiting patient outcomes.