7.1.462. neuro_deconvolve.pyΒΆ

Link to classic view

Generate a script that would apply 3dTfitter to deconvolve an MRI signal (BOLD response curve) into a neuro response curve.

Required parameters include an input dataset, a script name and an output prefix.

  1. deconvolve 3 seed time series

    The errts time series might be applied to the model, while the all_runs and fitts and for evaluation, along with the re-convolved time series generated by the script.

    Temporal partitioning is on the todo list.

    neuro_deconvolve.py

    -infiles seed.all_runs.1D seed.errts.1D seed.fitts.1D

-tr 2.0 -tr_nup 20 -kernel BLOCK
-script script.neuro.txt

old examples:

old 1. 3d+time example

neuro_deconvolve.py
-input run1+orig -script script.neuro
-mask_dset automask+orig
-prefix neuro_resp

old 2. 1D example

neuro_deconvolve.py
-input epi_data.1D -tr 2.0 -script script.1d -prefix neuro.1D
-help : display this help
-hist : display the modification history
-show_valid_opts : display all valid options (short format)
-ver : display the version number

-input INPUT_DATASET : set the data to deconvolve

e.g. -input epi_data.1D

-prefix PREFIX : set the prefix for output filenames

e.g. -prefix neuro_resp

–> might create: neuro_resp+orig.HEAD/.BRIK

-script SCRIPT : specify the name of the output script

e.g. -script neuro.script

-kernel KERNEL : set the response kernel

default: -kernel GAM

-kernel_file FILENAME : set the filename to store the kernel in

default: -kernel_file resp_kernel.1D
  • This data should be at the upsampled TR.

    See -tr_nup.

-mask_dset DSET : set a mask dataset for 3dTfitter to use

e.g. -mask_dset automask+orig

-old : make old-style script

Make pre-2015.02.24 script for 1D case.

-tr TR : set the scanner TR

e.g. -tr 2.0

The TR is needed for 1D formatted input files. It is not needed for AFNI 3d+time datasets, since the TR is in the file.

-tr_nup NUP : upsample factor for TR

e.g. -tr_nup 25

Deconvolution is generally done on an upsampled TR, which allows for sub-TR events and more accurate deconvolution. NUP should be the number of pieces each original TR is divided into. For example, to upsample a TR of 2.0 to one of 0.1, use NUP = 20.

TR must be an integral multiple of TR_UP.

-verb LEVEL : set the verbose level

e.g. -verb 2
  • R Reynolds June 12, 2008

Table Of Contents

This Page