MATLAB Answers

Unsure how to separate 4-D matrix data to perform correlations

7 views (last 30 days)
Sawyer on 11 Dec 2018
Commented: Sawyer on 13 Dec 2018
I have a file containing two different variables 1.REF and 2.TIME.
The dimension of the REF variable is 30*30*14*60, where there are 30 voxels in X direction, 30 voxels in Y direction and 14 voxels in Z direction and there are 60 timepoints (Volumes acquired). Variable TIME has a reference time series.
I need to write a program to perform the following:
1.Correlate Reference time series with each and every voxel time series and create a new 3 dimensional data
2.Find out the voxels with maximum correlation value
I've never worked with a 4-D matrix before so am unsure how to break up the REF variable so I can perform the correlations. Can someone please help me?


Sign in to comment.

Accepted Answer

Mark Sherstan
Mark Sherstan on 12 Dec 2018
This is common in image processing and you can use a similar process to extract your features. Using the image attachted if you run the following commands:
I = imread('Voxels.png');
>> 916 800 3
You can see the matrix is 916x800x3. Here the image is 916x800 pixels. The image is made out of red, green, and blue pixels of different ratios to make a specified color (hence the 3). So if I only wanted to extract the red components I would run:
red = I(:,:,1);
Similarly I could put 2 and 3 to get blue and green components respectively.
If the image was a gif it would be 916x800x3xN where N would be a frame like in a video. If I wanted the last red frame I could run:
red = I(:,:,1,end);
So in your case simply to extract all the features you could run a loop like this:
for ii = 1:60
for jj = 1:14
You might be better off to store the results in a cell array but it all depends on how you plan on analyzing your results (so ill leave that up to you). Hope this helped.

  1 Comment

Sign in to comment.

More Answers (0)




Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!