These scripts provide visualizations of brain scan data in MATLAB.
We analyze both fMRI [functional magnetic resonance imaging] data and utilize DTI [diffusion tensor imaging] methods.
current dataset: https://openneuro.org/datasets/ds000114/versions/1.0.2
paper: https://pmc.ncbi.nlm.nih.gov/articles/PMC3641991/pdf/2047-217X-2-6.pdf
downloading the data: Use the shell script ds000114-1.0.2.sh which includes curl commands. Change the permissions to make it executable:
chmod +x ds000114-1.0.2.sh
Then run
./ds000114-1.0.2.sh
to download the dataset which may take a few minutes.
fMRI animations:
- script =
ds000114_basic_animations_statistics.m - loads both functional (func) and anatomical data (anat)
- produces three animations as .gif files [temporal functional, spatial functional, spatial anatomical]
- computes basic statistics
- outputs a time series for a given voxel
DTI visualizations:
- guide: https://www.diffusion-imaging.com/
- script =
ds000114_dti_examples.m - computes four DTI values [AD, RD, MD, FA] for fixed voxel in
ds000114-1.0.2 - computes [AD, RD, MD, FA] for across slices
- displays and saves resulting 2d images
activation examples
- script =
ds000114_activation_examples.m - set paths for the dataset and GitHub project folder
- define canonical HRF (double gamma) for modeling brain response
- loop over subjects to load fMRI data and event timing
- build design matrix by convolving condition onsets with HRF
- fit voxelwise GLM to estimate beta maps for each condition
- save PNGs showing all conditions for each subject and the group average
paths: Ensure your path is set appropriately (change the username). Place the data downloaded from the shell script in the GitHub repository root directory.
- data downloaded from shell script:
'/Users/<your-username-here>/Documents/GitHub/neuroimage-analysis/ds000114-1.0.2'
datasets:
- https://community.ukbiobank.ac.uk/hc/en-gb/articles/24618819821981-Imaging-Data
- https://www.humanconnectome.org/study/hcp-young-adult
- https://openneuro.org
future brain imaging methods:
- MRA [magnetic resonance angiography]
future topics:
- multi-modal studies [combining different data types; genetic, questionaire data, GWAS etc.]
- neurodegenerative bases for disorders such as bipolar, schizophrenia [e.g. myelination]
articles: