Skip to content. | Skip to navigation

Sections
Personal tools
You are here: Home → Forum → Using data sets → vim-1:natural stimulus
 

vim-1:natural stimulus

Up to Using data sets

vim-1:natural stimulus

Posted by Dai Zhang at April 10. 2013
I am trying to use your natural stimulus to build a fMRI experiment. You provide the each stimulus images as a 500x500 matrix in .mat formt. However the value in the matrix is not grey value (0-255). I want to know how to transform the matrix to a grey value matrix.
For most matrixs, the maximum  value was 0.4645 and the minimum value was -0.5355. I also noticed that  for each image the background pixel was set to 0. You said that the background value was set to mean luminance across photographs, and a linear transformation was implement between ogrinal pixel value and the value after transformation. Could I match 0.4645 to 255 and -0.5355 to 0?
 
 

regards,

dai.

Re: vim-1:natural stimulus

Posted by Mark Lescroart at April 10. 2013

Yes, that is the correct transform.

Re: vim-1:natural stimulus

Posted by Maryam Gholami at April 23. 2014

hello, I have a question about presented natural images to subjects.

if you are using mind reading data set: data set includes a ( 25000*1750) matrix in which there are  25000 voxels  at rows and 1750 voxel BOLD responses at columns. each column represents the BOLD responses of 25000 voxels when a natural image is presented to subject.   there are 1750  natural images as stimulus, so there are 1750 different voxel patterns. if I want to classify voxel BOLD responses, what features can be used. I think the type of images that is shown to subject  can be used to distinguish voxels BOLD responses throughout the watching that type if images.but we do not have any categories for natural images. we only have 1750 images.   can you suggest me  a way to have limited patterns instead of 1750  voxel patterns?

Thanks, 

Mary

Re: vim-1:natural stimulus

Posted by Mark Lescroart at May 02. 2014

Hi Mary, 

As I said in the other thread, it's up to you to come up with a set of categories into which to divide the stimuli. You could, for example, label all the images that have faces or people in them as one category, and all the images that don't have people as another category. You can label anything you like. And as you can perhaps see, that could be a lot of work. But it is beyond the scope of our data sharing to define all the possible analyses you might want to do with our data.

Cheers,

Mark

Powered by Ploneboard
Document Actions