Automatic rejector in Data acquisition using Igor NIDAQ and NeuroMatic

Hi Pros,

I am building a data acquisition system and I find Igor NIDAQ and NeuroMatic pretty handy.

Problem:
I collect electrophysiological signals from Human subjects and they are not anethesized during recording. Therefore, I have electronic noise from body movements, blinking, etc. Previously, we use LabVIEW to collect data and we can set a range of signal amplitudes (microV) and reject the signals out of bounds, which are usually a result of body movements and blinking.
When I say reject, I mean signals out of bounds are displayed but not averaged. One single run with big noise in episodic recording mode can disrupt the average signal waveform dramatically, which obscure real-time judgement. I understand, it is no sweat to delete those runs with big noise and re-average after recording. But, is there a way to avoid saving those data in the first place during the experiment? Can it be a Igor Macro or Procedure file run in real-time in recording, automatically measures the delta-Peak value on y-axis (amplitude in microV), then reject if this delta-Peak is larger than the preset value?

How I can realize this function in Igor NIDAQ and NeuroMatic?

Thanks so much!

Eddie