Dwarf mongoose gannet optimization algorithm-enabled deep neuro-fuzzy network for detection of shockable ventricular cardiac arrhythmias

Abstract

One of the main sources of the Sudden Cardiac Death (SCD) is termed as Fatal arrhythmia. The electric shock treatment retrieves the regular electrical and mechanical functions of the heartbeat by controlling Ventricular fibrillation (VF) and Ventricular Tachycardia (VT). Shockable arrhythmia is easily controlled by providing electrical shock treatments. On the other hand, non-shockable arrhythmia is not controlled by electrical shock treatment. It is a very complex task to accurately discriminate these two kinds of arrhythmia using human assessment of Electrocardiogram (ECG) signals within a limited span and there may be a chance to occur faults during manual inspection. An accurate ECG diagnosis is very significant as it saves the life of the patient in advance by delivering proper therapy. To address this emerging problem, an automated model using the proposed Dwarf Mongoose Gannet Optimization Algorithm-Deep Neuro-Fuzzy Network (DMGOA-DNFN) is invented for detecting the shockable ventricular cardiac arrhythmias (SVCA). The classification is performed effectively utilizing DNFN and the weight of this classifier is optimally adjusted employing a newly developed algorithm named DMGOA, which is a consolidation of Dwarf Mongoose Optimization (DMO) and Gannet Optimization Algorithm (GOA). The proposed DMGOA-DNFN has surpassed other classical models with respect to accuracy of 93.2%, sensitivity of 95.8%, and specificity of 91.7%.