BIDS App containing an instance of the SPM12 software.
Extensive documentation can be found in the SPM manual.
To launch an instance of the container and analyse some data in BIDS format, type:
$ docker run bids/spm bids_dir output_dir level [--participant_label PARTICIPANT_LABEL [PARTICIPANT_LABEL ...]] [--config CFG_FILE]
For example, to run an analysis in participant
level mode, type:
$ docker run -ti --rm \
-v /tmp:/tmp \
-v /var/tmp:/var/tmp \
-v /path/to/local/bids/input/dataset/:/data \
-v /path/to/local/output/:/output \
bids/spm \
/data /output participant --participant_label 01
For example, to run an analysis in group
level mode with a user-defined pipeline, type:
$ docker run -ti --rm \
-v /tmp:/tmp \
-v /var/tmp:/var/tmp \
-v /path/to/local/bids/input/dataset/:/data \
-v /path/to/local/output/:/output \
-v /path/to/local/cfg/:/cfg \
bids/spm \
/data /output group --config /cfg/my_pipeline_group.m
To build the container, type:
$ docker build -t <yourhandle>/spm12 .
The configuration file is a MATLAB script detailing the analysis pipeline to be executed. Two struct variables, BIDS
and BIDS_App
are available from within the script, containing details from the command line and the BIDS-formatted dataset. In particular, the BIDS
structure can be queried using the spm_BIDS()
function. The default configuration files for first and second level analyses are pipeline_participant.m
and pipeline_group.m
. A template for a single configuration file for all levels could be as follow:
if strcmp(BIDS_App.level,'participant')
% First level analysis
% Get T1-weighted image filename for given subject:
% a = spm_BIDS(BIDS,'data', 'modality','anat', 'type','T1w');
% ...
else
% Second level analysis
% The name of the directory containing first level outputs is stored in:
% BIDS_App.outdir
% ...
end
If you have a specific problem with the SPM BIDS App, please open an issue on GitHub.
If your issue concerns SPM more generally, please use the SPM mailing list.
Please refer to:
@Book{spm,
editor = {K.J. Friston and J. Ashburner and S.J. Kiebel and T.E. Nichols and W.D. Penny},
title = {Statistical Parametric Mapping: The Analysis of Functional Brain Images},
publisher = {Academic Press},
year = {2007},
url = {http://store.elsevier.com/product.jsp?isbn=9780123725608}
}