Executive Summary is intended as the last stage in any DCAN fMRI pipeline. It is intended to show some key files to allow a quick QC of the image processing results of a single session of a single subject. This program was designed along with the DCAN fMRI pipelines, and depends upon their output.
This is not a stand-alone BIDS App.
The ExecutiveSummary.py
script runs in 2 steps: preprocessing and layout
of the HTML page. The preprocessing step makes an executivesummary
directory where the HTML file will eventually be written. It also makes an
img
subdirectory. Output from the preprocssing step is put into the
img
directory. For example, the preprocessor slices some of the BIDS input
(nii.gz
) files into .png
files in img
. The layout step writes the HTML
file. The ExecutiveSummary.py
script also uses some files produced by the
DCANBoldProcessing stage. As of this writing, some files stored in
img
do not have BIDS names.
You can move the Executive Summary output, to another directory (or device), but
it must be moved as a package. That is, the HTML must be in the same location as
the img
directory so it can find its images. Best to move the entire
executivesummary
directory as a unit.
- Imaging Software Packages Required:
- fsl v5.10 or later
- workbench v1.3.2 or later
- Environment:
- python 3.7.x
- argparse
- PIL (Python Image Library)
- Executive Summary is run as a stage of the DCAN fMRI pipelines.
After all of the stages of the pipeline, through DCAN BOLD processing have been run:
- Executive Summary can be run by passing the option
--stage ExecutiveSummary
as an option torun.py
. ExecutiveSummary.py
can be run from the command line as below:
usage: ExecutiveSummary [-h] --output-dir FILES_PATH [--bids-input FUNC_PATH]
--participant-label PARTICIPANT_LABEL
[--session-id SESSION_ID]
[--dcan-summary DCAN_SUMMARY] [--atlas ATLAS_PATH]
[--version] [--layout-only]
Builds the layout for the Executive Summary of the bids-formatted output from
the DCAN-Labs fMRI pipelines.
optional arguments:
-h, --help show this help message and exit
--output-dir FILES_PATH, -o FILES_PATH
path to the output files directory for all
intermediate and output files from the pipeline. Path
should end with "files".
--bids-input FUNC_PATH, -i FUNC_PATH
path to the bids dataset that was used as task input
to the pipeline. Path should end with "func"
--participant-label PARTICIPANT_LABEL, -p PARTICIPANT_LABEL
participant label, not including "sub-".
--session-id SESSION_ID, -s SESSION_ID
filter input dataset by session id. Default is all ids
found under each subject output directory(s). A
session id does not include "ses-"
--dcan-summary DCAN_SUMMARY, -d DCAN_SUMMARY
Optional. Expects the name of the subdirectory used
for the summary data. Default:
summary_DCANBOLDProc_v4.0.0
--atlas ATLAS_PATH, -a ATLAS_PATH
Optional. Expects the path to the atlas to register to
the images. Default:
templates/MNI_T1_1mm_brain.nii.gz.
--version, -v show program's version number and exit
--layout-only Can be specified for subjects that have been run
through the executivesummary preprocessor, so the
image data is ready. This calls only the
layout_builder to get the latest layout.
executivesummary/img
subdirectory containing:- Structurally registered image slices for each task acquisition.
- Concatenated grayordinates plots, pre-regression and post-regression, for entire run and for individual series.
- T1 and T2 .png files: images of each resting-state volume with orthogonal slice-positions.
executivesummary/executive_summary_sub-<label>.html
: a dashboard for cursory quality assurance.- BrainSprite viewer with navigable 3-D images.
- Carousels (aka sliders) to view T1 pngs, T2 pngs, and registered images.
- v2.1.0: Rearranged layout to add subcorticals.
- v2.0.0: Complete rewrite and new API. Handles task- data with different naming conventions. New features include carousels. NOTE: API has changed. This version is NOT backward compatible.
- v1.5.0: Support for ABCD, monkey, and infant images, ica flag, pulling TE and TI from DICOMs
- v1.3.0: add list-mode support! (supply a list of processed paths)
- v1.2.3: floating points now have only 2 decimal places
- v1.2.2: SBRef data can be found elsewhere when we do not have Raw