Main Article Content
Arrhythmia is one among the leading cardiovascular disease (CVDs), which is responsible for sudden loss of life among
the cardiac patients. In the past few years, a tremendous growth in the discipline of IoT is witnessed, which contributes a lot in the healthcare system as it enables continuous monitoring of the patients yet there is a need for an advanced automatic monitoring system for the classification of cardiac arrhythmia. The traditional methods experience a great deal of disadvantages in terms of the classification accuracy, which is addressed through proposing an optimized deep convolutional neural network (Deep CNN) for IoT cardiac arrhythmia classification. The proposed technique assure ceaseless healthcare monitoring of a patient as it employs the IoT networks to collect the Electrocardiograph (ECG) signal, which is considered as a significant modality for arrhythmia classification.
The proposed optimized deep CNN is developed through the consolidation of the rider optimization algorithm (ROA) in the deep CNN classifier for tuning the hyper-parameters. The proposed model is evaluated with MIT–BIH dataset and the outcomes are analysed with the existing methods in order to reveal the efficacy of the proposed optimized deep CNN technique. The exploration of the classification methods based on accuracy, sensitivity and specificity reveals that the proposed method acquires an effective specificity with 98.8%, accuracy with 98.7% and sensitivity with 98.9%
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.