Hello,

I'd like to create an impulse response model. My functional data are not continuous (e.g. sparse) and thus it is not appropriate to insert stimulus timing files. For example, my timing files begin at the beginning of the experiment, but the functional data are not continuous. So if I use TENT or CSPLIN then it would take the timing file and generate my impulses accordingly, except that it wouldn't be in the same temporal domain as the functional data (because the fMRI data isn't continuous).

So I have my own impulse response matrix that I would like to provide where columns refer to an event type and rows refer to a specific volume.

For example, here is a sample of 6 rows:

1 0 0 0 0 0 0 0 0 0 0 0
0 1 0 0 0 0 0 0 0 0 0 0
0 0 1 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 1 0 0
0 0 0 0 0 0 0 0 0 0 1 0
0 0 0 0 0 0 0 0 0 0 0 1

Here is how I am getting the matrix into 3dDeconvolve.

3dDeconvolve 												                \
			-input $runs -jobs 24										\
			-polort A 	 										                \
			-ortvec MoPar_demean.1D 	demean 						        \
			-ortvec MoPar_derv.1D 	        derv 						        \
			-ortvec FIR.1D 		        ir	# impulse response matrix	        \
			-censor censor.1D 						                                        \
			-num_glt 1 											        \
			-gltsym 'SYM: ir[9..11] -ir[0..2]'  -glt_label 1 sound-nosound 			\

Obviously there are more deconvolve options for statistics and such, I just wanted to provide how I'm setting up the relevant parts.
I want to be able to submit some of the impulse response columns to a glt test. In the above example, lets say that I want to do a t test between columns 9 through 11 minus columns 0 through 2. How would I specify this?

If I specify it in the above way I receive an error:

** ERROR: row #1 of matrix 'SYM: ir{9]}' is all zero!
** ERROR: -gltsym errors immediately above from file 'SYM: ir{9]}'
** FATAL ERROR: Can't continue after the above -gltsym problems!

My rationale for this specification is because this is how I'd do it in TENT or CSPLIN. Thank you for your time.

Dustin