BrainIAK Tutorials: Condensed Neuroimaging Datasets

BrainIAK tutorials provide a collection of pre-processed neuroimaging datasets, condensed from original studies to reduce file size. The collection includes datasets such as VDC, Ninety Six, Face-scene, Latatt, Pieman2, Raider, and Sherlock_processed, paired with specific tutorials for practical application. These datasets are ready for use in teaching and demonstrating methods for functional magnetic resonance imaging (fMRI) data analysis.

Use Cases

• Tutorial-based learning of fMRI data handling using the simulated 02-data-handling dataset.
• Within-subject searchlight analysis demonstration using the VDC dataset.
• Multi-voxel pattern analysis (MVPA) practice with the Face-scene dataset.
• Temporal functional connectivity exploration using the Pieman2 and Raider datasets.
• Naturalistic narrative processing analysis with the Sherlock_processed dataset.

Strengths

• Datasets are pre-processed and ready for immediate use, reducing setup time.
• Collection is condensed from original studies, with a total unzipped size of 18GB, making it manageable for tutorial purposes.
• Includes both real and simulated data, covering multiple cognitive neuroscience paradigms.

Limitations

• Specific column names, row counts, and individual dataset sizes are not provided in the source metadata.
• The last updated date is unknown, which may affect confidence in the currency of the data processing pipelines.
• The dataset is a condensed version with a reduced number of subjects, which limits its utility for full-scale research replication.

Provenance

Source

Compiled by Manoj Kumar from Princeton University for BrainIAK tutorials, sourcing data from published neuroimaging studies.

Collection Method

Datasets were condensed from original published studies by reducing the number of subjects.

Time Range

Studies referenced span from 2008 to 2017.