Differences

This shows you the differences between two versions of the page.

--- gsr_data_analysis [2010/06/17 21:03] – fixed an error (cell size) jochen
+++ gsr_data_analysis [2010/06/18 23:22] (current) – changed onsets to be in ms jochen
@@ Line 22: / Line 22: @@
 <code matlab GSR_data_readacq.m>% use file selector of xff class
-gsr = xff('*.acq', 'Please select the GSR-ACQ file to work with...');
+gsr = xff('*.acq', 'Please select the GSR-ACQ file to work with...');</code>
-% the actual data will then either be in
+To use a variable in a MAT file, simply use the ''load MATFILE'' call in Matlab:
-data = gsr.RawData(CHANNEL, :)';
+<code matlab GSR_data_loadmat.m>% load a mat file (e.g. an ACQ->MAT converted file)
+load HPS1344_session1_GSR.mat;
-% or in
-data = gsr.Channel(CHANNEL).Data;</code>
+% create new NTT (used for methods on data!)
+ntt = xff('new:ntt');
-To use a variable in a MAT file, simply use the ''load MATFILE'' call in Matlab.
+% store data from mat file in ntt
+ntt.Data = data;</code>
 Or if you require to load the data from a text-based (e.g. log) file, you can use the following code:
@@ Line 52: / Line 54: @@
 {{:gsr_example_5secsraw.png|10 seconds of raw GSR data recording}}
-As you can see, the very small signal variations are normally quite uninteresting (as the skin response only very sluggishly follows the stimulation/emotional state of the subject). Hence, it is usually a good idea to resample the signal to a more suitable resolution. To keep some of the finer details (and being able to determine latency values with a reasonable resolution), for this example I resampled the data to 100Hz using the [[resampleaa]] function (which works a little better than using just ''downsample''):
+As you can see, the very small signal variations are normally quite uninteresting (as the skin response only very sluggishly follows the stimulation/emotional state of the subject). Hence, it is usually a good idea to resample the signal to a more suitable resolution. To keep some of the finer details (and being able to determine latency values with a reasonable resolution), for this example I resampled the data to 100Hz using the [[acq.Resample|ACQ::Resample]] or [[ntt.Resample|NTT::Resample]] method (internally using the [[resampleaa]] function which works a little better than using just ''downsample''):
 <code matlab GSR_resampledplot.m>% resampling the data
-rdata = resampleaa(data, 5);
+% the first channel is actual physiological data and needs hi-quality
+% resampling (cubic/gaussian average), the stim channels 2:9
+% only need indexing into the correct temporal samples
+gsr.Resample(100, struct('cubic', [true, false(1, 8)]));
 % plotting same time index with stronger line
-set(plot(rdata(50400:51399)), 'LineWidth', 2);</code>
+data = gsr.ChannelData(1);
+set(plot(data(50400:51399)), 'LineWidth', 2);</code>
 {{:gsr_example_5secsresampled.png|100Hz-resampled GSR data}}
 ===== (Pre-) Filtering data =====
-While a lot of the very high-frequency noise is (naturally) gone after the resampling, there is still a lot of signal variation that is (very likely) unrelated to any task/stimulation the subject could have reacted to. Further the signal pre-processing, I then applied a difference (first discreet derivative, ''diff'')-filter that removes artifacts in signals which are unlikely to be in the desired spectrum using the [[prefilter]] function:
+While a lot of the very high-frequency noise is (naturally) gone after the resampling, there is still a lot of signal variation that is (very likely) unrelated to any task/stimulation the subject could have reacted to. Further the signal pre-processing, I then applied a difference (first discreet derivative, ''diff'')-filter that removes artifacts in signals which are unlikely to be in the desired spectrum using the [[acq.Filter|ACQ::Filter]]/[[ntt.Filter|NTT::Filter]] method (which makes use of the [[prefilter]] function):
-<code matlab GSR_prefilter.m>% simply pass to prefilter
+<code matlab GSR_prefilter.m>% call obj.Filter with default parameters for first channel
-fdata = prefilter(rdata);
+gsr.Filter(1);
 % plotting same time index with stronger line
-set(plot(fdata(50400:51399)), 'LineWidth', 2);</code>
+data = gsr.ChannelData(1);
+set(plot(data(50400:51399)), 'LineWidth', 2);</code>
 {{:gsr_example_5secsfiltered.png|filtered GSR data}}
@@ Line 82: / Line 89: @@
 <code matlab GSR_acq_stimonsets.m>% all stimuli onsets (use 1 + to cope for the fact that the
 % first value in diff rather codes the second sample in the data!
-all_onsets = 1 + find(diff(gsr.RawData(8, :)') > 0);
+all_onsets = 1 + find(diff(gsr.ChannelData(8)) > 0);
 % now split into three conditions
-neu = all_onsets(gsr.RawData(2, all_onsets) > 0);
+neu = 0.01 * all_onsets(gsr.ChannelData(2, all_onsets) > 0);
-neg = all_onsets(gsr.RawData(3, all_onsets) > 0);
+neg = 0.01 * all_onsets(gsr.ChannelData(3, all_onsets) > 0);
-rea = all_onsets(gsr.RawData(4, all_onsets) > 0);
+rea = 0.01 * all_onsets(gsr.ChannelData(4, all_onsets) > 0);</code>
-% cope for resampling from 500Hz -> 100Hz !!
-neu = round(0.2 .* neu);
-neg = round(0.2 .* neg);
-rea = round(0.2 .* rea);</code>
 As window length for the inspection of the GSR data, we choose 15 seconds, so in the resolution of the resampled data that means 1,500 samples.
+In our second example, the data only contains the onsets (not the condition information). This has to be established separately (which can be even done **after** extracting the peak-to-peak values, thus allowing for a blind-analysis of data, as the examiner doesn't know which onset belongs to which specific condition):
+<code matlab GSR_plain_onsets.m>% loading the data
+% (in this case a MAT file written by Acqknowledge for ACQ Version >= 4)
+% which is in 200Hz sampling frequency with
+% GSR = channel 1 and
+% ECG = channel 2
+load 6701.acq.mat
+% the stimulus channel in this example is channel 3,
+% coded with 0 Volts for no stimulus and 5V for stimulus
+% thus, to find the time points (samples) on which a
+% stimulus occurs, we look for spots in this channel
+% where the value changes from below 2.5V to above 2.5V
+stimactive = (data(:, 3) > 2.5);
+onsets = find(~stimactive(2:end) & stimactive(1:end-1));
+% creating the necessary information for plotcurves
+% the onset is shifted by 200 samples (1 second) to show
+% a window of [-1 .. 14] (in seconds)
+c = cell(numel(onsets), 2);
+for cc = 1:numel(onsets)
+    c{cc,1} = sprintf('Onset %d, t=%.3f', cc, onsets(cc)/200);
+    c{cc, 2} = [1, onsets(cc)-200, onsets(cc)+2800];
+end
+% calling plotcurves
+plotcurves(data, struct('curves', {c}, 'freq', 200));</code>
 ===== Peak-to-peak analysis =====
@@ Line 105: / Line 136: @@
 reacurves = cell(numel(rea), 2);
 for sc = 1:numel(neu)
-    neucurves{sc, 1} = sprintf('LookNeu %02d at t=%.3fs', sc, 0.01 * neu(sc));
+    neucurves{sc, 1} = sprintf('LookNeu %02d at t=%.3fs', sc, neu(sc));
-    neucurves{sc, 2} = [gsrchannel, neu(sc), neu(sc) + 1499];
+    neucurves{sc, 2} = [gsrchannel, neu(sc) - 2, neu(sc) + 14];
 end
 for sc = 1:numel(neg)
-    negcurves{sc, 1} = sprintf('LookNeg %02d at t=%.3fs', sc, 0.01 * neg(sc));
+    negcurves{sc, 1} = sprintf('LookNeg %02d at t=%.3fs', sc, neg(sc));
-    negcurves{sc, 2} = [gsrchannel, neg(sc), neg(sc) + 1499];
+    negcurves{sc, 2} = [gsrchannel, neg(sc) - 2, neg(sc) + 14];
 end
 for sc = 1:numel(rea)
-    reacurves{sc, 1} = sprintf('ReaNeg %02d at t=%.3fs', sc, 0.01 * rea(sc));
+    reacurves{sc, 1} = sprintf('ReaNeg %02d at t=%.3fs', sc, rea(sc));
-    reacurves{sc, 2} = [gsrchannel, rea(sc), rea(sc) + 1499];
+    reacurves{sc, 2} = [gsrchannel, rea(sc) - 2, rea(sc) + 14];
 end
+% spots
+spot = {[neu; neg; rea]};
+spotnames = {'Stim onset'};
 % combine curves
@@ Line 127: / Line 162: @@
 % use plot curves
-plotcurves(gsr, struct('curves', {curves}, 'sets', {sets}));</code>
+plotcurves(gsr, struct( ...
+    'curves',    {curves}, ...
+    'sets',      {sets}, ...
+    'spot',      {spot}, ...
+    'spotnames', {spotnames}));</code>