NIH Blueprint: The Human Connectome Project

Summary of 7T MR imaging protocols

Resting state fMRI, movie-watching fMRI, retinotopy fMRI, and dMRI acquisitions were collected over 4 total imaging sessions, each approximately 1.25 hours in duration. Each session started with a resting-state fMRI acquisition of approximately 16 minutes, followed by 2 movie-watching fMRI (sessions 1 and 4), 2 retinotopy fMRI (session 2), or 2 diffusion acquisitions (session 3) of varying durations (see below).

Vitamin E capsule on right side. A capsule of vitamin E was taped to the subject’s right temple in every scan session, to enable definitive determination of the right side in the image data.

The following provides basic parameters for the main scan types in each session, and pertinent details about each session. A more complete set of imaging parameters can be found in the 7T protocol exports from the scanner, available for all 7T subjects as the 7T MRI Session Summary CSVs (184 subjects) in ConnectomeDB on the S1200 Project page as a “Quick Download”. FOV positioning in all runs was handled in an automated manner using the Siemens AutoAlign feature. Eyetracking data was collected for all fMRI scans and is available in the unprocessed 7T data packages for each scan type.

Resting-state fMRI (rfMRI)

rfMRI data were acquired in four runs of approximately 16 minutes each, each at the beginning of each of the four 7T imaging sessions, with eyes open with relaxed fixation on a projected bright cross-hair on a dark background (and presented in a darkened room). Among the 4 rfMRI scans across the 4 sessions, oblique axial acquisitions alternated between phase encoding in a posterior-to-anterior (PA) direction in runs 1 and 3, and an anterior-to-posterior (AP) phase encoding direction in runs 2 and 4.

Resting state and all 7T fMRI images were collected with the following parameters:

Parameter Value
Sequence Gradient-echo EPI
TR 1000 ms
TE 22.2 ms
flip angle 45 deg
FOV 208 x 208 mm (RO x PE)
Matrix 130 x 130 (RO x PE)
Slice thickness 1.6 mm; 85 slices; 1.6 mm isotropic voxels
Multiband factor 5
Image Acceleration factor (iPAT) 2
Partial Fourier (pF) sampling 7/8
Echo spacing 0.64 ms
BW 1924 Hz/Px
Condition Runs Frames (TRs) Per Run Run Duration (min:sec)
REST (Resting-state) 4 900 16:00

Movie-watching fMRI

Following completion of rfMRI in the first and fourth 7T fMRI scanning sessions, subjects were scanned in 2 runs while watching short (ranging from 1 to 4.3 minutes in length) independent film and Hollywood movie exerpts concatenated into .mp4 files of 11.9-13.7 minutes total length, including a common Vimeo repeat validation clip.

The movie-watching tfMRI data were acquired with the same EPI pulse sequence parameters as R-fMRI, except for the run duration, which was dictated by the stimuli presented.

The table below lists the movie-watching scans collected, with “CC” denoting movies composed of clips of freely available independent films under Creative Commons licensing and “HO denoting movies composted of clips from Hollywood movies as prepared and published by Cutting et al. 2012.

.mp4 files used as stimulus for each scan are named identically to the scan name. There are four movie scans in the protocol comprising just over one hour of total tfMRI scan time, with two movie scans collected in 7T session 1 and two more movie scans collected in session 4. Two runs were acquired with posterior-to-anterior (PA) phase encoding and the remaining two runs were acquired with anterior-to-posterior (AP) phase encoding so that bias may be cancelled out in analyses across multiple runs.

We used two versions of the four .mp4 movie files compiled as stimuli, as described in detail in Movie watching stimuli and experimental details below. The “Pre_20140821_version” was used for subjects collected from the beginning of Phase 2 scanning through August 21, 2014 and the "Post_20140821_version", or v2, was used for subjects from August 22, 2014 to the end of 7T data collection.

Movie-watching scan Runs Phase Encoding Frames Per Run Run Duration (min:sec)
Pre_20140821_version: 4
7T_MOVIE1_CC1 AP 921 15:21
7T_MOVIE2_HO1 PA 918 15:18
7T_MOVIE3_CC2 PA 915 15:15
7T_MOVIE4_HO2 AP 901 15:01
Post_20140821_version: 4
7T_MOVIE1_CC1_v2 AP 921 15:21
7T_MOVIE2_HO1_v2 PA 918 15:18
7T_MOVIE3_CC2_v2 PA 915 15:15
7T_MOVIE4_HO2_v2 AP 901 15:01

Retinotopy task fMRI

Following completion of rfMRI in the second 7T fMRI scanning session, subjects were scanned in 6 runs while watching retinotopic stimuli videos, each 5 minutes in length, for a total of 30 minutes of scan time.

Retinotopy Scan Retinotopy Stimulus Phase Encoding Run Duration (min:sec)
RET1 RETCCW (Counter Clockwise sweep) AP 5:00
RET2 RETCW (Clockwise sweep) PA 5:00
RET3 RETEXP (Expanding Circle) AP 5:00
RET4 RETCON (Contracting Circle) PA 5:00
RET5 RETBAR1 (Multidirection Bar sweeps ) AP 5:00
RET6 RETBAR2 (same stimulus as RETBAR1) PA 5:00

The experimental viewing distance for all runs was 101.5 cm, with a stimulus diameter of 28.5 cm and stimulus size of 16.0°.

Diffusion imaging (dMRI)

Parameter Value
Sequence Spin-echo EPI
TR 7000 ms
TE 71.2 ms
flip angle 90 deg
refocusing flip angle 180 deg
FOV 210x210 (RO x PE)
matrix 200x200 (RO x PE)
slice thickness 1.05 mm, 132 slices, 1.05 mm isotropic voxels
Multiband factor 2
Image Acceleration factor (iPAT) 3
Echo spacing 0.82 ms
BW 1388 Hz/Px
Phase partial Fourier 6/8
b-values 1000, 2000 s/mm2

Other details: Diffusion gradients are monopolar. Oblique axial acquisitions alternate between right-to-left and left-to-right phase encoding directions in consecutive runs. Image reconstruction uses SENSE1 multi-channel (Sotiropoulos et al., 2013).

A full dMRI session includes 4 runs (each approximately 9 minutes and 50 seconds), representing 2 different gradient tables, with each table acquired once with anterior-to-posterior and posterior-to-anterior phase encoding polarities, respectively. Each gradient table includes approximately 65 diffusion weighting directions plus 6 b=0 acquisitions interspersed throughout each run. Diffusion weighting consisted of 2 shells of b=1000 and 2000 s/mm2 interspersed with an approximately equal number of acquisitions on each shell within each run. The diffusion directions were obtained using a toolbox available from Emmanuel Caruyer that returns uniformly distributed directions in multiple q-space shells. The directions are optimized so that every subset of the first M directions is also isotropic. References and the toolbox can be found at: http://www.emmanuelcaruyer.com/q-space-sampling.php.

Full scanning protocols (PDF)

Complete scanning protocols for each 7T imaging session are available as separate PDF outputs from the scanner:

Standard 7T MR session structure

Each subject was scanned by MRI in four regular sessions (~4.5 hours total), with the following shorthand labels:

Day 1:

  • Session 1 (mov1_7T): One 16-min rfMRI scan (REST1_PA) and two movie tfMRI scans (MOVIE1_AP and MOVIE2_PA)
  • Session 2 (ret_7T): One 16-min rfMRI scan (REST2_AP) and six retinotopy tfMRI scans (RET1_AP, RET2_PA, RET3_AP, RET4_PA, RET5_AP, and RET6_PA one RL and one LR scan for each task)

Day 2:

  • Session 3: (diff_7T): One 16-min rfMRI scan (REST3_PA) and four dMRI scans (DWI_AP_dir71, DWI_AP_dir71, DWI_AP_dir71, and DWI_AP_dir71)
  • Session 4: (mov2_7T): One 16-min rfMRI scan (REST4_AP) and two movie tfMRI scans (MOVIE3_PA and MOVIE4_AP)

Note: The temporal order of actual scan acquisitions for each subject is reflected in the scan number in ConnectomeDB (and also as an explicit ScanOrder variable).

In some cases, scans were acquired in an extra session (coded session 5 or higher, as appropriate). This was done if the data quality was inadequate in the regular scan or if technical problems prevented a full set of acquisitions in the regular session.