This build context will create an image built to the Flywheel Gear Specification, which can execute Freesurfer's recon-all
(v6.0.1) within Flywheel, or locally.
- You MUST read and agree to the license agreement and register with MGH before you use the software.
- Once you get your license you can edit the
manifest.json
file to include your license details before you build the container. Without a license the execution of the code will fail. - This image is built with the Matlab MCRv84 included. The MCR is required to run the optional Hippocampal Subfields and Brainstem Structures processing (see
manifest.json
). - The resulting image is ~12GB and builds in ~15min.
Configuration for running the algorithm (and adding the license) are defined within manifest.json
. The options available, along with their defaults, are described in the manifest.json
file.
If you would like to use specific options in a local run of this gear you can modify the default
key for each option, which will be used during the local run - executed when executed with Docker.
This Gear is designed to run within Flywheel, however you can run this Gear locally. To run recon-all
from this image you can do the following:
# Note that the `recon-all` command is omitted as it is called from the `Entrypoint`.
docker run --rm -ti \
-v </path/to/input/data>:/input/flywheel/v0/input/anatomical \
-v </path/for/output/data>:/output \
scitran/freesurfer-recon-all:<version-tag>
- You must mount the directory (using the
-v
flag) which contains your anatomical data (nifti or dicoms) in the container at/input/flywheel/v0/input/anatomical
and also mount the directory where you want your output data stored at/output
, see the example above. - Configuration options (including the license key) must be set in the
manifest.json
file before building the container.