This dataset contains the Midnight Scanning Club (MSC) data, a dataset focused on the precise characterization of ten individual subjects via collection of large amounts of per-individual data. Each subject underwent twelve separate two-hour scanning sessions. In the first two sessions we collected four T1 images, four T2 images, four MR angiograms, and eight MR venograms. In the last ten sessions we collected five hours of resting-state fMRI data and over five and a half hours of task fMRI data across three different tasks. Participants also underwent extensive neuropsychological testing. These raw data are all provided here.

In addition to the raw data, we also provide several derivatives processed using both a volumetric (Talaraich-space) and a surface-based (fs_LR_32k space) pipeline. Volumetric derivatives include cross-session-average T1 and T2 images linearly registered to atlas (Talaraich) space; as well as resting-state data from each scanning session that has been fully preprocessed, motion-censored, and confound-regressed. Surface pipeline derivatives include cortical surfaces which were segmented from the T1 scans using freesurfer, hand-edited, and registered to fs_LR atlas space; resting-state data from each scanning session that has been fully preprocessed, motion-censored, and confound-regressed in CIFTI format (cortical: fs_LR32k; subcortical: Talaraich); cortical parcellations estimated from the resting-state data; vertex-wise whole-brain networks estimated from the resting-state data; task timecourses in CIFTI (cortical: fs_LR32k; subcortical: Talaraich) space; a selection of task contrast images in CIFTI (cortical: fs_LR32k; subcortical: Talaraich) space; myelin maps estimated from the T1- and T2-weighted anatomical scans; and matrices describing the physical geodesic/euclidean distances between every two points in the cifti image.

Details of this dataset are more fully described in Gordon EM, Laumann TO, Gilmore AW, Newbold DJ, Greene DJ, Berg JJ, Ortega M, Hoyt-Drazen C, Gratton C, Sun H, et al. (2017). Precision Functional Mapping of Individual Human Brains. Neuron 95, 791–807. This manuscript should be cited when publishing work using this data.

Experiment

Three healthy adult participants wore a cast covering the entire right upper extremity for two weeks. They were scanned every day for 6-9 weeks. Scans included 42-64 daily 30-minute resting-state functional MRI scans before, during and after casting. Participants later underwent 12-24 additional scans as part of a control experiment. In all, we collected 27-43 hours of resting-state functional MRI data in each individual.

Participants also performed a block-design movement task (right hand, left hand, right foot, left foot, tongue) for 8 minutes each night prior to casting.

Details of this dataset are described in Newbold et al., Plasticity and Spontaneous Activity Pulses in Disused Human Brain Circuits, Neuron (2020), https://doi.org/10.1016/j.neuron.2020.05.007. This manuscript should be cited whenever publishing work using this dataset.

Raw data

Sessions are grouped into 5 conditions. 3 conditions (pre, cast, post) correspond to the original experiment. 2 conditions (on, off) correspond to a control experiment in which participants wore a removable cast during scanning (on sessions) but were not casted during daily life.

Two participants were also studied in a previous experiment, the Midnight Scan Club (MSC) experiment (Gordon et al, 2017, https://openneuro.org/datasets/ds000224). sub-cast1 was sub-MSC02. sub-cast2 was sub-MSC06. Carried forward many of the methods from the MSC experiment to the current study.

MSC participants were scanned using a 3T Siemens Trio MRI scanner. BOLD data were acquired at a spatial resolution of 4mm, single-band, with a TR of 2.2s. We used identical sequences for sub-cast1 during the original cast experiment (but not during the later control experiment).

After running sub-cast1, a new MRI scanner became available. sub-cast2 and sub-cast3 were scanned on a 3T Siemens Prisma using new sequences. The updated scanner and sequences were also appleid to sub-cast1 during the later control experiment. BOLD data for these scans were acquired at a spatial resolution 2.4mm, multi-band 4, with a TR of 1.1s.

Pre-processed data

In addition to the raw BOLD and structural data we collected, we have also provided fully pre-processed rs-fMRI and task-fMRI data. Processing pipelines are described in Newbold et al, 2020 and all processing scripts are available on GitLab (https://gitlab.com/DosenbachGreene/cast-induced-plasticity). Processed data are provided in volume space as well as cifti space -- combined cortical surface data and sub-cortical/cerebellar volume data.

Surface projection followed methods described in Gordon et al, 2017. Derivative structural files needed for cifti creation (e.g. pial/white surfaces, subcortical masks) are provided for sub-cast2 and sub-cast3. Because sub-cast1 was scanned using the same scanner and sequences used for the MSC study, cortical projections for sub-cast1 used the projection files generated for sub-MSC02 (https://openneuro.org/datasets/ds000224, derivatives/surface_pipeline/sub-MSC02/fs_LR_Talairach/).

Parcellations

Surface parcellations for sub-cast3 were created using methods described by Gordon et al (2017). Corresponding parcellations for sub-cast1 and sub-cast2 can be found in the MSC dataset (https://openneuro.org/datasets/ds000224, derivatives/surface_pipeline/sub-{subject}/surface_parcellation).