Efficient method to perform a simple linear regression across dimensions

I have a 4D array of neuroimaging data (nSubj*x*y*z), and a 2D array of covariates (nSubj*1).

What is an efficient way to do a separate regression for every voxel (i.e. value of x, y, and z) to the array of covariates, resulting in a 3D array of regression values (size=x*y*z)?

Thanks!

Answer 1

[Star Strider]

The best I can do is to direct you to the general section in the Statistics Toolbox documentation on Linear Regression and let you explore. I cannot suggest anything specific, because I’m not certain what you’re doing.