Tags: example statistics freq cluster

Apply clusterrandanalysis on TFRs of power that were computed with BESA

% this is the list of BESA data files in one condition
filename_ld = {
  'AM_LD_ERA.tfc'
  'CGi_LD_ERA.tfc'
  'DK_LD_ERA.tfc'
  'ES_LD_ERA.tfc'
  'HK_LD_ERA.tfc'
  'JS_LD_ERA.tfc'
  'LJ_LD_ERA.tfc'
  'LPO_LD_ERA.tfc'
  'PP_LD_ERA.tfc'
  'SH_LD_ERA.tfc'
  };

% this is the list of BESA data files in the other condition
filename_sd = {
  'AM_SD_ERA.tfc'
  'CGi_SD_ERA.tfc'
  'DK_SD_ERA.tfc'
  'ES_SD_ERA.tfc'
  'HK_SD_ERA.tfc'
  'JS_SD_ERA.tfc'
  'LJ_SD_ERA.tfc'
  'LPO_SD_ERA.tfc'
  'PP_SD_ERA.tfc'
  'SH_SD_ERA.tfc'
  };

nsubj = length(filename_ld);

% collect all single subject data in a convenient cell-array
for i=1:nsubj
  ld{i} = besa2fieldtrip(filename_ld{i});
  sd{i} = besa2fieldtrip(filename_sd{i});
end

% this is needed for the channel labels in the data from Peter Praamstra
% MATLAB is case sensitive and we want the channel and electrode names to match
for i=1:nsubj
  for j=1:length(ld{i}.label)
    if ld{i}.label{j}(end)=='H'
      ld{i}.label{j}(end)='h';
    end
  end
  for j=1:length(sd{i}.label)
    if sd{i}.label{j}(end)=='H'
      sd{i}.label{j}(end)='h';
    end
  end
end

% load a set of electrodes (these are on a unit sphere)
% note, this will be different for your own data
load('elec128.mat');
% scale the electrodes to a realistic head size (in cm)
elec.pnt = 10*elec.pnt;

% compute the grand average for both conditions
cfg = [];
ldavg = ft_freqgrandaverage(cfg, ld{:});
sdavg = ft_freqgrandaverage(cfg, sd{:});

% make a dummy structure with the difference between ld and sd
avgdif = ldavg;
avgdif.powspctrm = ldavg.powspctrm - sdavg.powspctrm;

% make some figures
cfg = [];
cfg.layout = elec;
figure; ft_multiplotTFR(cfg, sdavg);
figure; ft_multiplotTFR(cfg, ldavg);
figure; ft_multiplotTFR(cfg, avgdif);

% perform the statistical test using randomization and a clustering approach
% using the ft_freqstatistics function
cfg = [];
cfg.elec             = elec;
cfg.neighbourdist    = 4;
cfg.latency          = 'all';
cfg.frequency        = 'all';
cfg.channel          = 'EEG1010' % see FT_CHANNELSELECTION
cfg.avgovertime      = 'no';
cfg.avgoverfreq      = 'no';
cfg.avgoverchan      = 'no';
cfg.statistic        = 'ft_statfun_depsamplesT';
cfg.numrandomization = 200;
cfg.correctm         = 'cluster';
cfg.method           = 'montecarlo';
cfg.design           = [
  1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10     % subject number
  1 1 1 1 1 1 1 1 1 1  2 2 2 2 2 2 2 2 2 2  ];  % condition number
cfg.uvar = 1;                                   % "subject" is unit of observation
cfg.ivar = 2;                                   % "condition" is the independent variable
stat = ft_freqstatistics(cfg, ld{:}, sd{:});