Reward Documentation (2020)

Last updated: July 1, 2020
Written by Anna Xu

Reward is a dataset comprised of 6 projects collected over the span of years, totaling 509 participants (note that some participants were present in multiple projects). These are the 6 projects with the scans available in them:

Imaging & Task Data Status (Flywheel)

Scan data & associated task behavioral data


All scan data found on XNAT and associated task behavioral data found on XNAT or CfN has been uploaded onto Flywheel in the project Reward2018. Scan data was de-identified by uploading with the command fw import dicom ${dicomUploadPath} mcieslak "Reward2018" --subject ${bblid} --session ${subproject} -y --profile "dicom_config.yaml" in the command line and with the configuration file dicom_config.yaml found on Flywheel in Reward2018 > Information.

The following procedure checked to make sure all expected scans on XNAT were accounted for on Flywheel:

Scanid for each participant has also been noted in the custom info for each subject/session.


These scans have also been fw’heudiconv-ed using these heuristics and the command fw-heudiconv-curate --project "Reward2018" --heuristic "${heuristicFile}" --session "${subproject}" --subject "${bblid}".

You can also use the script to mass run fw-heudiconv for a select few participants.


Some scans have been processed with an old version of fMRIPrep but all scans may need to be fMRIPrep’d again with the latest version of fMRIPrep. details the previous process of running fMRIPrep.

Task files

Files documenting all nifti scans and their associated task behavioral data can be found on Flywheel in the file fwScansTask_20200624.csv. This file is located in Reward2018 > Information and contains the following variables:

Another file, missingTaskFiles_20200625.csv documents a summary of the missing task files. This file is currently on the reward_data_mgmt slack channel (unsure if this is also on the Reward2018 project on Flywheel due to inability to check Flywheel before my departure). This file contains the following:

If needed, goes through the process of generating a large spreadsheet of all files present in the Reward2018 project. To generate these associated files, use the script scanTaskDocumentation.R which sources code from classify_scans.R and classify_task.R.

Other data

Demographics, Medication Status, Diagnosis

Demographics, medication status, diagnosis, and effort data are currently with Dan (not uploaded anywhere). Missing data has been previously reported in the reward_data_mgmt slack channel.

Aylin’s Monster

Aylin’s monster scripts can be found in the Reward GitHub repo by clicking here. A wiki page of old documentation for Aylin’s monster can be found here.