Within Ponemah, Noise Detection is performed in two ways:  RMS and Min Heart Rate. 

When noise is detected, bad data marks will be placed on the noisy data.  Noise Detection by RMS and Bad Data Threshold Attribute.  RMS for each cycle is calculated to determine how noisy the cycle is. The QRS complex, P wave, and T wave are excluded from RMS calculation from the last R to the current R region. The algorithm is: first, excluding the first 10% and last 10% of data, which include the QRS complex from the R to R region. Then, look for the largest derivative values and exclude 5% of data both before and after the largest derivative position. Regard this as a P (or T) wave. Again, we look for the largest derivative value within the remaining region and exclude 5% of data both before and after the largest derivative position. Regard this as another T (or P) wave. Then, 60% of data remains to calculate RMS. On the following graph, derivative values over the green regions are used to calculate RMS.

If the RMS calculated is larger than Bad Data Threshold specified, a bad data mark will be put and no ECG mark will be put between the bad data mark. The default Bad Data Threshold is 100.