How to correct NeuroScore from scoring an epoch as SWS in spite of large EMG activity

Background:

NeuroScore sometimes scores an epoch as SWS in spite of large EMG activity.

Workaround:

When the activity signal is greater than 0, the epoch is likely to be scored as Active Wake.

The idea is to add a certain number (+6 in the example below) to the Activity signal whenever the EMG RMS signal is above a certain threshold.

This way, even if the Activity signal is 0, it will become 6 if there is significant EMG activity, and the epoch will be scored as AW.

In the example below, epoch # 1611 is scored as SWS, whereas it's clearly AW:

Increasing the Delta threshold could help : this epoch may get scored as AW. However, some other epochs might then be scored incorrectly. So it may not be the best solution.

Here's what can be tried:

Looking at the EMG RMS signal, determine a threshold above which you want the epoch to be scored as Active Wake, no matter what the other signals look like. In this example, let's say the threshold is 30 µV (or 0.00003 V).
Right-click on the EMG RMS signal and select Add Derived Signal:
Select Transform and enter the Expression as shown below (note that the 0.00003 I've highlighted corresponds to the threshold determined above):
You will get a signal that has a '0' value, unless the EMG RMS value is above the threshold, in which case the value will be -6:
On the right hand side of the window, in the 'Signal and Data' tab, click 'New Rereferenced Signal', and make the following selections:

You will obtain a new Activity signal, which is like the old one with +6 when the EMG RMS is above 30 µV.
Run the Rodent Sleep Scoring detector, and select 'NewActivity' when prompted to select the input signal for Activity:
Now epoch # 1611 is scored as AW:

Note that the following epoch (# 1612) is now scored as AW as well.

This is because the epoch prior to # 1612 is #1611 which just changed from SWS to AW, so #1612 changes as well due to the following matrix: