MRI dataset description

The dataset currently consist of a hanful of participants, being scanned as part of the course. The dataset will increase over the years.

Where to find the data on the server

The dataset is stored in /home/lab1user/SHARED_labdata/PSYC724_LAB-DATA/ This is a shared space available from each machine.

That folder contains 4 subfolders

  • 01_sourcedata contains the files coming directly from the scanner in dicom format.

  • 02_rawdata contains the files converted in nifti format. This copy of the dataset will serve as a backup in case the files are corrupted during the preprocessing.

  • 03_derivatives will contain all the files generated by the preprocessing and analyses steps

  • 04_scripts will contain the matlab script written to perform the analyses.

You should start by getting your local copy of the 03_derivatives folder. Open a file explorer and in /home/lab1user/STUDENT-DATA/, create a folder called PSYC724_LAB-DATA/ where you can copy 03_derivatives. Do not copy the source and raw data.

Data structure

The scanner produces images in a format called dicom (.dcm). This is a standard medical format but statistical packages only deal with a different format, called nifti (.nii).

The dataset has a standardized structure: Each subject folder contains an anatomical directory and a functional directory labeled anat and func, and these in turn contain the anatomical and functional images collected during the experiment. The files are named according to BIDS, or Brain Imaging Data Structure, which makes it easy to organize and analyze and share your data.

../_images/02_dataset_structure.png

The func directory contains functional data - in this case, five runs of functional data. The files are in 4D, they contains times series of 3D images. The json file contains information about the image acquisition from the scanner. The files “events.tsv” contain onsets, or timestamps of which condition happened at what time. They have been created from the logfiles generated by the stimulus presentation software, contained in beh You can open these as a text file or as a spreadsheet. We will use these later when we build our General Linear Model.

The fmap directory contains filedmap images that can be used to correct for distortions.

Functional image files contain 170 contiguous echo planar imaging (EPI) whole-brainfunctional volumes (Multiband factor=3; TR=1500 ms; TE=30 ms; flip angle=72, 72 slices, matrix=108*108; FOV=216 mm; acquisition voxel size=2x2x2mm) during each of the five task runs. A high-resolution T1-weighted anatomical image was also acquired using a magnetization prepared gradient echo sequence (MPRAGE, TR=2100 ms; TE=2.26 ms; flip angle=8; 176 slices, FOV=256 mm).