Prediction of cardiac rhythm based on heart rate variability features in sinus tachycardia conditions

The variability observed in cardiac rhythm measurements makes classification and prediction of Sinus Tachycardia (ST) ECG signals challenging. The study aims to predict ST volunteer's heart rate by designing and developing a Long Short Term Memory (LSTM) based Recurrent Neural Network (RNN) model. The Heart Rate Variability (HRV) features comprising the time and frequency domain were considered as the input feature set for the designed and developed LSTM-RNN model. The proposed LSTM-RNN model works on the principle of multivariate time series forecasting. The LSTM-RNN model's training accuracy is 96.72%, and the validation accuracy is 85.21%. The performance of the LSTM model works well for predicting time series of unknown duration. The time gap in the input time series data does not affect the LSTM model, giving an advantage over other time series prediction models.