This article describes how to obtain the Power Spectral Density of each sleep stage using NeuroScore and Excel.

A rodent sleep recording will be used in the following example. Automatic sleep scoring was performed with NeuroScore.

​​


Create a Signal Grid, with the following columns :
- Add Column > Time Stamp
- Add Column > Sleep Scoring
- Add Column > Signal > EEG > DFT Power Band > Band Range = 0 to 0.4 Hz
- Add Column > Signal > EEG > DFT Power Band > Band Range = 0.5 to 0.9 Hz
- Add Column > Signal > EEG > DFT Power Band > Band Range = 1 to 1.4 Hz
- Add Column > Signal > EEG > DFT Power Band > Band Range = 1.5 to 1.9 Hz
- ... [repeat as many times as needed]


​

Then comes a very important step : create an artifact detector (amplitude detector) to insert a marker called "Signal-Invalid" whenever the signal is above a certain threshold (4 mV in this example) :

​​​​​

And run the detector on the EEG signal (the amplitude detector can be run on 4 different signals, just skip the last 3 windows) :

​​

The Power Spectrum of the artifacted epoch has huge Power values. They must be eliminated before averaging is performed.

Observe that the lines where an artifact was detected have been removed from the Signal Grid :

​


Then Export the grid to Excel, and average the lines based on the Sleep Stage.

In this example, we got the following Power Density Spectrums :

Slow-Wave-Sleep :

​​

Paradoxical Sleep :

​

Wake :

​