The Neurophysiological Biomarker Toolbox (NBT)

## Computing amplitudes in the classical frequency bands

EEG signals are often categorized in the classical frequency bands delta (1-3 Hz), theta (3-7 Hz),alpha (8-13 Hz), beta (16-36 Hz) and gamma (36-45 Hz). The analysis function nbt_doAmplitude computes the integrated amplitude for each of these frequency bands, for each channel, and for six subregions of the brain (in case of 129 channel EEG data), for which we computed the mean over channels within the subregions. Moreover, the normalized amplitudes, i.e., the amplitude in a frequency band divided by the integrated amplitude over all frequency bands, are also computed. In case of 129-channel EEG data, the results can be visualized and stored in PDFs (see below for example plots). For each frequency band, the results are stored in an amplitude biomarker object (see nbt_Amplitude for the definition of these objects). The analysis function nbt_runAmplitude stores these biomarker objects in the analysis file.

If you want to compute amplitudes for only one NBT Signal: in the NBT GUI (see The NBT GUI and loading a NBT Signal ) select Compute biomarkers | For current NBT signal | Amplitudes.

Alternatively, via the command line:

nbt_runAmplitude(Signal,Info,SaveDirectory)

Inputs:
Signal = NBT Signal matrix.
Info = NBT Info object.
SaveDirectory = a full path to the folder you want to save the analysis file.

For example:

nbt_runAmplitude(Signal,SignalInfo,'B:/NBT files')

If you want to compute amplitudes for a bunch of NBT files in a folder: in the GUI, select Compute biomarkers | For multiple NBT signals | Amplitudes.

Alternatively, via the command line:

nbt_NBTcompute(@nbt_runAmplitude,SignalName,LoadDirectory)

Inputs:
SignalName = string with the name of the signal you want to use, if you do not have more signals in your signal file choose 'Signal'.
LoadDirectory = string with the name of the directory where your NBT files are.

For example:

nbt_NBTcompute(@nbt_runAmplitude,'Signal','B:/NBT files')

Example of the plots that are generated, in case of 129-channel EEG data: