How to save indices of column in timeseries2matrix function

[prateeksasan1]

Hi,

I am trying to convert a 4-d image to a matrix using a mask using timeseries2matrix. I am also interested in noting what part of the mask corresponds to what column. Is it possible?

For e.g 4-d image is of dimension 90100801200 (1200 is the time point)
and i have a 3-d binary mask (dim = 90100*80) with sum(mask) = 150

The output of timeseries2matrix would be a matrix of size 1200 * 150.

I want to know which indices of the mask the columns correspond to. i.e 1st column is mask index (1, 4, 5) , 2nd is (1, 4, 6) etc.

Is there an easy way to do this. I see in your code that you subset the image which makes it a vector and then turn it into a matrix but couldnt figure out how to get the indices

[stnava]

use antsGetNeighborhood* functionality with radius=0 .... read the documentation brian

…

On Mon, Dec 7, 2020 at 3:05 PM Prateek Sasan ***@***.***> wrote: Hi, I am trying to convert a 4-d image to a matrix using a mask using timeseries2matrix. I am also interested in noting what part of the mask corresponds to what column. Is it possible? For e.g 4-d image is of dimension 90*100*80 *1200 (1200 is the time point) and i have a 3-d binary mask (dim = 90*100*80) with sum(mask) = 150 The output of timeseries2matrix would be a matrix of size 1200 * 150. I want to know which indices of the mask the columns correspond to. i.e 1st column is mask index (1, 4, 5) , 2nd is (1, 4, 6) etc. Is there an easy way to do this. I see in your code that you subset the image which makes it a vector and then turn it into a matrix but couldnt figure out how to get the indices — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#335>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AACPE7TE42WHHP6GI2OKTETSTUYQLANCNFSM4UQ6NM6Q> .

[prateeksasan1]

Thanks. I was not aware of this function.

Using getNeighborhoodInMask with spatial.info = TRUE, the 2nd matrix in the list gave me the indices. Awesome. Thanks a lot

Prateek Sasan