An Integrated Approach for Early Risk Detection of Sudden Cardiac Death Using Machine Learning Approach
Abstract
To sustain the survival of heart patients, early diagnosis of sudden cardiac death might be encouraged. In this paper we examined an algorithm for predicting and controlling sudden cardiac death by evaluating the heart rate variability signal utilizing methods of classical and temporal frequency. At first, just before the cardiac death case, one minute of ECG signals are eliminated and used to assess pulse variance in heart rate (HRV).We proposed optimum early detection of sudden cardiac (OEDSC) scheme to boost classification efficiency. Next, we implemented the Improved Whale Optimization Algorithm (IWOA) to detect the sudden cardiac disease and normal from the provided dataset for experienced feature selection process and Random forest classifier. The proposed method was assessed using an ECG database. The proposed method was implemented using a random forest classifier and had a valid detection rate (accuracy) of 91.16 percent for classical features and TF method, respectively.
Keywords: Heart rate variability (HRV), Optimal early detection of sudden cardiac (OEDSC), Improved Whale Optimization algorithm (IWOA), Time-frequency (TF).