Difference between revisions of "Specific research subjects"

From biophysics
Jump to navigation Jump to search
(Created page with "<p class="clearnone"> A tutorial, for students and coworkers: a recipe to calibrate, detect, and analyze saccades</p> <hr id="system-readmore" /> <p class="clearnone">Easily...")
 
(Replaced content with "==Saccades== Saccade Cookbook")
Tag: Replaced
Line 1: Line 1:
<p class="clearnone"> A tutorial, for students and coworkers: a recipe to calibrate, detect, and analyze saccades</p>
+
==Saccades==
<hr id="system-readmore" />
 
  
 
+
[[Saccade Cookbook]]
<p class="clearnone">Easily calibrate, detect, and analyze <a href="http://en.wikipedia.org/wiki/Saccade" target="_blank">saccades</a> with <a href="&lt;p class=&quot;clearnone&quot;&gt; &lt;a href=&quot;https://gitlab.science.ru.nl/marcw/biofysica&quot; target=&quot;_blank&quot;&gt;PANDA&lt;/a&gt; tutorial, for students and coworkers: a recipe to calibrate, detect, and analyze saccades&lt;/p&gt;">PandA</a>. For a more detailed description, visit the other <a href="http://www.neural-code.com/index.php/panda/action/saccade">saccade</a> sections, starting with the <a href="http://www.neural-code.com/index.php/panda/action/87-saccade/70">introduction</a>.</p>
 
 
 
<h2>Calibration</h2>
 
<p>First you have to calibrate (see also saccade <a href="http://www.neural-code.com/index.php/tutorials/action/saccade/55-saccade-calibration">calibration</a>)
 
  by training an artificial neural netwok to learn the relationship between measured voltages and head orientation:
 
<pre xml:lang="matlab">pa_calibrate</pre>
 
  <p>(choose the calibration dat-file in the pop-up menu, usually ending in 0000).
 
  Then you can calibrate (and low-pass filter) the data-files, by using the button in the pa_calibrate interface
 
  (this will calibrate ALL dat-files with the .net-file you got from the pa_calibrate-procedure). </p>
 
 
 
<h2>Detection</h2>
 
<p>Then you have to detect (see also saccade <a href="http://www.neural-code.com/index.php/tutorials/action/saccade/56-saccade-detection">detection</a>)
 
with:
 
<pre xml:lang="matlab"> pa_sacdet</pre>
 
<p>(choose an experimental hv-file, usually ending with numbers 0001 and up).  </p>
 
<h2>Parameters</h2>
 
<p>Obtain all movement parameters, such as end-point location, reaction time, peak velocity. </p>
 
<pre xml:lang="matlab"> pa_sac2mat</pre>
 
  (these will be saved in a mat-file)</p>
 
 
 
<h2>Analysis</h2>
 
<p>Finally, you can start some high-level analysis (see also saccade <a href="http://www.neural-code.com/index.php/tutorials/action/saccade/54-saccade-analysis">analysis</a>)
 
  After all these obligatory steps, you can actually start analyzing the data. This usually involves custom-made analysis functions, suited for your experiment. First you have to load the data:
 
<pre xml:lang="matlab"> load MW-RG-2011-03-02-0001</pre>
 
<p>You will now have a Sac- and a Stim-matrix. The Sac-matrix contains all the movement parameters, the Stim-matrix all the stimulus parameters. You can combine them into a single matrix, containing in each column a relevant parameter, and each row containing a single saccade.</p>
 
<pre xml:lang="matlab"> SupSac = pa_supersac(Sac,Stim,XX,YY);</pre>
 
<p>Note that the XX and YY should be numbers describing the type of stimulus, and the number of that type of stimulus in the trial, e.g.
 
<pre xml:lang="matlab"> SupSac = pa_supersac(Sac,Stim,1,2);</pre>
 
<p>for the second (YY=2) stimulus of type 1 (XX=1, usually an LED) in a trial, or</p>
 
<pre xml:lang="matlab"> SupSac = pa_supersac(Sac,Stim,2,1);</pre>
 
<p>for the first (YY=1) stimulus of type 2 (XX=2, usually a sound) in the trial (you have to check what number is assigned to which stimulus type, by typing:
 
<pre xml:lang="matlab"> help pa_readcsv</pre>
 
in the trial information of the LOG-matrix, line 5).</p>
 
<p>To plot target location versus response location :</p>
 
<pre xml:lang="matlab"> plot(SupSac(:,23),SupSac(:,8),’k.’);</pre>
 
<p>or</p>
 
<pre xml:lang="matlab">pa_plotloc(SupSac);</pre>
 

Revision as of 13:40, 22 January 2024