Note: This dataset is the combination of four related datasets that were originally hosted on OpenfMRI.org: ds000113, ds000113b, ds000113c and ds000113d. The combined dataset is now in BIDS format and is simply referred to as ds000113 on OpenNeuro.org.

For more information about the project visit: http://studyforrest.org

This dataset contains high-resolution functional magnetic resonance (fMRI) data from 20 participants recorded at high field strength (7 Tesla) during prolonged stimulation with an auditory feature film ("Forrest Gump''). In addition, a comprehensive set of auxiliary data (T1w, T2w, DTI, susceptibility-weighted image, angiography) as well as measurements to assess technical and physiological noise components have been acquired. An initial analysis confirms that these data can be used to study common and idiosyncratic brain response pattern to complex auditory stimulation. Among the potential uses of this dataset is the study of auditory attention and cognition, language and music perception as well as social perception. The auxiliary measurements enable a large variety of additional analysis strategies that relate functional response patterns to structural properties of the brain. Alongside the acquired data, we provide source code and detailed information on all employed procedures — from stimulus creation to data analysis. (https://www.nature.com/articles/sdata20143)

The dataset also contains data from the same twenty participants while being repeatedly stimulated with a total of 25 music clips, with and without speech content, from five different genres using a slow event-related paradigm. It also includes raw fMRI data, as well as pre-computed structural alignments for within-subject and group analysis.

Additionally, seven of the twenty subjects participated in another study: empirical ultra high-field fMRI data recorded at four spatial resolutions (0.8 mm, 1.4 mm, 2 mm, and 3 mm isotropic voxel size) for orientation decoding in visual cortex — in order to test hypotheses on the strength and spatial scale of orientation discriminating signals. (https://www.sciencedirect.com/science/article/pii/S2352340917302056)

Finally, there are additional acquisitions for fifteen of the the twenty participants: retinotopic mapping, a localizer paradigm for higher visual areas (FFA, EBA, PPA), and another 2 hour movie recording with 3T full-brain BOLD fMRI with simultaneous 1000 Hz eyetracking.

For more information about the project visit: http://studyforrest.org

Dataset content overview

Stimulus material and protocol descriptions

./sourcedata/acquisition_protocols/04-sT1W_3D_TFE_TR2300_TI900_0_7iso_FS.txt ./sourcedata/acquisition_protocols/05-sT2W_3D_TSE_32chSHC_0_7iso.txt ./sourcedata/acquisition_protocols/06-VEN_BOLD_HR_32chSHC.txt ./sourcedata/acquisition_protocols/07-DTI_high_2iso.txt ./sourcedata/acquisition_protocols/08-field_map.txt Philips-specific MRI acquisition parameters dumps (plain text) for structural MRI (T1w, T2w, SWI, DTI, fieldmap -- in this order)

./sourcedata/acquisition_protocols/task01_fmri_session1.pdf ./sourcedata/acquisition_protocols/task01_fmri_session2.pdf ./sourcedata/acquisition_protocols/angio_session.pdf Siemens-specific MRI acquisition parameters dumps (PDF format) for functional MRI and angiography.

./stimuli/annotations/german_audio_description.csv

Audio-description transcript

This transcript contains all information on the audio-movie content that cannot be inferred from the DVD release — in a plain text, comma-separated-value table. Start and end time stamp, as well as the spoken text are provided for each continuous audio description segment.

./stimuli/annotations/scenes.csv

Movie scenes

A plain text, comma-separated-value table with start and end time for all 198 scenes in the presented movie cut. In addition, each table row contains whether a scene takes place indoors or outdoors.

./stimuli/generate/generate_melt_cmds.py Python script to generate commands for stimuli generation

./stimuli/psychopy/buttons.csv ./stimuli/psychopy/forrest_gump.psyexp ./stimuli/psychopy/segment_cfg.csv Source code of the stimuli presentation in PsychoPy

Functional imaging - Forrest Gump Task

Prolonged quasi-natural auditory stimulation (Forrest Gump audio movie)

Eight approximately 15 min long recording runs, together comprising the entire duration of a two-hour presentation of an audio-only version of the Hollywood feature film "Forrest Gump" made for a visually impaired audience (German dubbing).

For each run, there are 4D volumetric images (160x160x36)in NIfTI format , one volume recorded every 2 s, obtain from a Siemens MR scanner at 7 Tesla using a T2*-weighted gradient-echo EPI sequence (1.4 mm isotropic voxel size). These images have partial brain coverage — centered on the auditory cortices in both brain hemispheres and include frontal and posterior portions of the brain. There is no coverage for the upper portion of the brain (e.g. large parts of motor and somato-sensory cortices).

Several flavors of raw and preprocessed data are available:

Raw BOLD functional MRI ~~~~~~~~~~~~~~~~~~~~~~~

These raw data suffer from severe geometric distortions.

Filename examples for subject 01 and run 01

./sub-01/ses-forrestgump/func/sub-01_ses-forrestgump_task-forrestgump_acq-raw_run-01_bold.nii.gz BOLD data

./sourcedata/dicominfo/sub-01/ses-forrestgump/func/sub-01_ses-forrestgump_task-forrestgump_acq-raw_run-01_bold_dicominfo.txt Image property dump from DICOM conversion

Raw BOLD functional MRI (with applied distortion correction) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Identical to raw BOLD data, but with a scanner-side correction for geometric distortions applied (also include correction for participant motion). These data are most suitable for analysis of individual brains.

Filename examples for subject 01 and run 01

./sub-01/ses-forrestgump/func/sub-01_ses-forrestgump_task-forrestgump_acq-dico_run-01_bold.nii.gz BOLD data

./derivatives/motion/sub-01/ses-forrestgump/func/sub-01_ses-forrestgump_task-forrestgump_acq-dico_run-01_moco_ref.nii.gz Reference volume used for motion correction. Only runs 1 and 5 (first runs in each session)

./sourcedata/dicominfo/sub-01/ses-forrestgump/func/sub-01_ses-forrestgump_task-forrestgump_acq-dico_run-01_bold_dicominfo.txt Image property dump from DICOM conversion

Raw BOLD functional MRI (linear anatomical alignment) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

These images are motion and distortion corrected and have been anatomically aligned to a BOLD group template image that was generated from the entire group of participants.

Alignment procedure was linear (image projection using an affine transformation). These data are most suitable for group-analyses and inter-individual comparisons.

Filename examples for subject 01 and run 01

./derivatives/linear_anatomical_alignment/sub-01/ses-forrestgump/func/sub-01_ses-forrestgump_task-forrestgump_rec-dico7Tad2grpbold7Tad_run-01_bold.nii.gz BOLD data

./derivatives/linear_anatomical_alignment/sub-01/ses-forrestgump/func/sub-01_ses-forrestgump_task-forrestgump_rec-dico7Tad2grpbold7TadBrainMask_run-01_bold.nii.gz Matching brain mask volume

./derivatives/linear_anatomical_alignment/sub-01/ses-forrestgump/func/sub-01_ses-forrestgump_task-forrestgump_rec-XFMdico7Tad2grpbold7Tad_run-01_bold.mat 4x4 affine transformation matrix (plain text format)

Raw BOLD functional MRI (non-linear anatomical alignment) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

These images are motion and distortion corrected and have been anatomically aligned to a BOLD group template image that was generated from the entire group of participants.

Alignment procedure was non-linear (image projection using an affine transformation with additional transformation by non-linear warpfields). These data are most suitable for group-analyses and inter-individual comparisons.

Filename examples for subject 01 and run 01

./derivatives/non-linear_anatomical_alignment/sub-01/ses-forrestgump/func/sub-01_ses-forrestgump_task-forrestgump_rec-dico7Tad2grpbold7TadNL_run-01_bold.nii.gz BOLD data

./derivatives/non-linear_anatomical_alignment/sub-01/ses-forrestgump/func/sub-01_ses-forrestgump_task-forrestgump_rec-dico7Tad2grpbold7TadBrainMaskNLBrainMask_run-01_bold.nii.gz Matching brain mask volume

./derivatives/non-linear_anatomical_alignment/sub-01/ses-forrestgump/func/sub-01_ses-forrestgump_task-forrestgump_rec-dico7Tad2grpbold7TadNLWarp_run-01_bold.nii.gz Warpfield (associated affine transformation is identical with "linear" alignment

Functional imaging - Auditory Perception Session

Participants were repeatedly stimulated with a total of 25 music clips, with and without speech content, from five different genres using a slow event-related paradigm.

Filename examples for subject 01 and run 01

./sub-01/ses-auditoryperception/func/sub-01_ses-auditoryperception_task-auditoryperception_run-01_bold.nii.gz ./sub-01/ses-auditoryperception/func/sub-01_ses-auditoryperception_task-auditoryperception_run-01_events.tsv

Functional imaging - Localizer Session

Filename examples for subject 01 and run

Extension of the dataset published in Hanke et al. (2014; doi:10.1038/sdata.2014.3) with additional acquisitions for 15 of the original 20 particpants. These additions include: retinotopic mapping, a localizer paradigm for higher visual areas (FFA, EBA, PPA), and another 2h movie recording with 3T full-brain BOLD fMRI with simultaneous 1000 Hz eyetracking. Only open-source software was employed in this study. We thank their respective authors for making it publicly available. Please follow good scientific practice by citing the most appropriate publication(s) describing the aspects of this datasets that were used in a study.

A manually denoised audio-visual movie watching fMRI dataset for the studyforrest project

Note: This dataset is compatible with the BIDS v1.3.0-dev as a standalone derivative dataset. Due to that the OpenNeuro.org does not support the BIDs v1.3.0 yet ('desc-

Source data

We here provide a denoised version of the studyforrest 3T movie-watching fMRI data. Specifically, the studyforrest 3T movie-watching fMRI dataset was hosted as ‘ses-movie’ under OpenNeuro ds000113 (https://openneuro.org/datasets/ds000113). Historically, the studyforrest 3T movie-watching fMRI data was hosted as ds000113d on OpenfMRI (https://legacy.openfmri.org/dataset/ds000113d/). It shall be noted that ‘ses-movie’ under OpenNeuro ds000113 and OpenfMRI ds000113d are the same data, just different in names.

For more information about the studyforrest project, please refer to the data papers of the studyforrest project and visit http://studyforrest.org.
Hanke, M. et al. A high-resolution 7-Tesla fMRI dataset from complex natural stimulation with an audio movie. Sci. Data 1, sdata20143 (2014).
Hanke, M. et al. A studyforrest extension, simultaneous fMRI and eye gaze recordings during prolonged natural stimulation. Sci. Data 3, 160092 (2016).

Pipeline description

A four-step procedure was used to denoise the movie-watching fMRI dataset from the studyforrest project, producing a denoised version of the data.

1. The data were preprocessed using FEAT in FSL v6.00 (https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FEAT) including motion correction, slice timing correction, brain extraction, high-pass temporal filtering (200s cutoff) and with or without spatial smoothing (i.e., FWHM= 5 mm or FWHM=0 mm).
2. ICA was performed with MELODIC v3.15 (https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/MELODIC).
3. IC classification was manually performed using melview (https://git.fmrib.ox.ac.uk/fsl/melview).
4. Artifact removal was performed using fsl_regfilt in FSL’s MELODIC v3.15.

All code for data denoising and technical validation can be found at https://github.com/xingyu-liu/studyforrest_denoise.

Dataset content overview

After the four-step denoising procedure, 4 kinds of data were produced for each run of each participant.

1. the denoised fMRI data
   ./sub-xx/ses-movie/func/sub-xx_ses-movie_task-movie-run-x_space-T1w_desc-{sm5,unsm}Denoised_bold.nii.gz
2. spatial maps of decomposed ICs
   ./sub-xx/ses-movie/func/sub-xx_ses-movie_task-movie-run-x_space-T1w_desc-{sm5,unsm}MELODIC_components.nii.gz
3. timeseries of decomposed ICs
   ./sub-xx/ses-movie/func/sub-xx_ses-movie_task-movie-run-x_space-T1w_desc-{sm5,unsm}MELODIC_mixing.tsv
4. Labels of decomposed ICs
   ./sub-xx/ses-movie/func/sub-xx_ses-movie_task-movie-run-x_space-T1w_desc-{sm5,unsm}MELODIC_decomposition.json

Collection description

Subject species

homo sapiens

Modality

fMRI-BOLD

Cognitive paradigm (task)

film viewing

Map type

Z

Study Forrest. Extensive functional brain imaging data from natural stimulation, a rich set of auxiliary data (such as structural brain scans, measurements of physiological, and technical confounds), as well as stimulus annotations.

This repository contains the fMRI data, annotations, analysis scripts to generate the results, and results in Häusler C.O. & Hanke M. (submitted) as Datalad datasets (https://github.com/datalad).

Support vector machine searchlight accuracy map. Radius of the spherical ROIs was two voxels (plus center). ROIs were spaced by two voxels. Accuracy scores mapped onto a voxels are the mean accuracy of all ROI analysis a voxel was included into.

Collection description

The twenty participants were repeatedly stimulated with a total of 25 music clips, with and without speech content, from five different genres using a slow event-related paradigm. The data release includes raw fMRI data, as well as precomputed structural alignments for within-subject and group analysis. In addition to fMRI, simultaneously recorded cardiac and respiratory traces, as well the complete implementation of the stimulation paradigm, including stimuli, are provided. An initial quality control analysis reveals distinguishable patterns of response to individual genres throughout a large expanse of areas known to be involved in auditory and speech processing.

Subject species

homo sapiens

Modality

fMRI-BOLD

Cognitive paradigm (task)

auditory scene perception

Map type

Other

This dataset contains key characteristics about the data described in the Data Descriptor A manually denoised audio-visual movie watching fMRI dataset for the studyforrest project. Contents:

    1. human readable metadata summary table in CSV format


    2. machine readable metadata file in JSON format 


      Versioning Note:Version 2 was generated when the metadata format was updated from JSON to JSON-LD. This was an automatic process that changed only the format, not the contents, of the metadata.

A collection of 30 brain maps. Each brain map is a 3D array of values representing properties of the brain at different locations.

Collection description

Faces and face related variables in studyForrest + music

Collection description

Faces and face related variables in studyForrest + music

Subject species

homo sapiens

Modality

fMRI-BOLD

Analysis level

group

Cognitive paradigm (task)

None / Other

Map type

T

A collection of 7 brain maps. Each brain map is a 3D array of values representing properties of the brain at different locations.

Collection description

Whole-brain activation maps for different type of transitions in naturalistic stimuli (using Studyforrest database)

Activation during location transitions, when the setting changed from the previous frame

Collection description

Whole-brain activation maps for different type of transitions in naturalistic stimuli (using Studyforrest database)

Subject species

homo sapiens

Modality

fMRI-BOLD

Analysis level

group

Cognitive paradigm (task)

movie watching task

Map type

T