Distribution of premature heartbeats

There are several methods to detect arrhythmia by analysis of the heartbeat rate. In practice, doctors follow a specific procedure to interpret if a particular RR interval as a premature heartbeat by analyzing the complex morphology of the ECG signal. We have tested several algorithms to determine premature heartbeats analyzing two neighboring RR intervals using different threshold values with a goal to obtain the smallest mistake. The results show that the best algorithm is based on calculating the arithmetic mean of the last 6 RR intervals and the RL6 ratio between the analyzed RR interval and the mean. The threshold value, which achieves a good estimation and classification precision has value of 0.16. This algorithm can be efficiently used as a fast pipelined algorithm in a mobile device saving its limited resources, such as battery and performance.