Within in Ponemah, Noise Detection is performed in two ways:  1.RMS, and 2. 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. QRS complex, P wave and T wave are excluded from RMS calculation from the last R to current R region. The algorithm is: first, exclude 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 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, have 60% data remain 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, bad data mark will be put and no ECG mark will be put between bad data mark. The default Bad Data Threshold is 100.