Global Field Power Calculation Difference

mcp0228 · September 23, 2022, 5:57am

Hello Brainstorm community,

The below diagram shows the result of using the CLUSTER function for POWER to generate the global field power of cortical auditory evoked potentials from 64 channels of electrodes.

I then enlarged the bottom green GFP trace from the top left figure to see if it looks exactly the same as the figure on the top right figure which is shown below.
I know that the y-axis scale is very different but I still thought that the general shape of the GFP should be very similar (if not identical) between the two figures, but they are not at the moment.
What might be the potential reasons?

Raymundo.Cassani · September 23, 2022, 6:22pm

Hi @mcp0228,

Indeed, they should be identical.

For the moment, the green GFP trace on top of the channel time series actually corresponds to sqrt(GFP)

github.com

brainstorm-tools/brainstorm3/blob/master/toolbox/gui/figure_timeseries.m#L3643


      
                      PlotHandles.hLineDecodingY = [];

                  end

                  % Create new lines

                  PlotHandles.hLineDecodingY = line([TimeVector(1), TimeVector(end)], [50 50], [0.5 0.5], ...

                      'LineWidth', 1, ...

                      'LineStyle', '--', ...

                      'Color',     .8*[1 1 1], ...

                      'Parent',    hAxes, ...

                      'Tag',       'LineDecoding50');

              end

          end

          

          % ===== DISPLAY GFP =====

          % GFP reference: https://link.springer.com/article/10.1007/BF01128870

          % If there are more than 5 channel

          if bst_get('DisplayGFP') && ~strcmpi(GlobalData.DataSet(iDS).Measures.DataType, 'stat') ...

                                   && (GlobalData.DataSet(iDS).Measures.NumberOfSamples > 2) && (size(F,1) > 5) ...

                                   && ~isempty(TsInfo.Modality) && ismember(TsInfo.Modality, {'EEG','MEG','EEG','SEEG'})

              GFP = std(F * fFactor, 1);

              PlotGFP(hAxes, TimeVector, GFP, TsInfo.FlipYAxis, isFastUpdate);

          end

We'll promptly solve this discrepancy.

Best,
Raymundo

Francois · September 24, 2022, 10:11am

After checking the literature, the most standard definition of the GFP is the standard deviation across sensors:

Brainstorm was computing a RMS measure instead (sqrt(sum(x.^2)), and then normalizing by the maximum for display (brainstorm3/figure_timeseries.m at master · brainstorm-tools/brainstorm3 · GitHub). The results is usually very similar or event identical as the STD because the mean of all the sensor values is most of the time close to zero (or IS equal to zero in the case of EEG with an average reference).

But to match the most common definition, I updated the code and documentation to use the STD instead:

However, the difference between GFP and POWER(cluster) you reports is normal: the former is the square root of the latter.