Detection and Classification of Arrhythmia Using Hybrid Deep Learning Model

Cardiovascular disorders, encompassing arrhythmias rhythms, represent noteworthy worldwide health issues that necessitate prompt identification and precise categorization to facilitate efficacious therapeutic measures. This study examines the status of the field's research at the moment, going into cardiac arrhythmias' physiological foundation, clinical ramifications, and diagnostic resources. The paper proposes a methodology that combines MobileNet V1 with Gated Recurrent Unit (GRU) for the detection and classification of Arrhythmia, which has achieved an accuracy of 99.4%, recall score of 0.98 and precision score of 0.99. The MIT-BIH Arrhythmia dataset, consisting of recordings by the BIH Arrhythmia Laboratory, is utilized for training and validation of the model. The findings demonstrate the potential of using the MobileNetV1+GRU model for detection and classification of different types of Arrhythmia, which can be applied for diagnosis in hospitals, clinics, emergency departments and screening programs.