Home

Menu

Loading wiki pages...

View
Wiki Version:
<p>This project investigates whether the way the brain encodes perceived actions allows it to simultaneously make automatic predictions about future actions.</p> <p>fMRI data consist of 17 participants who watched the first half of the first episode of the TV show <em>Sherlock</em>. These data are available <a href="https://dataspace.princeton.edu/jspui/handle/88435/dsp01nz8062179" rel="nofollow">here</a>. </p> <p>Actions in the Sherlock video were automatically annotated using a pre-trained temporal relation network, available <a href="https://github.com/metalbubble/TRN-pytorch" rel="nofollow">here</a>. Due to copyright restrictions, we do not post the video file publicly. If you wish to use this file for scientific purposes, please contact us or the authors of the original study in which the fMRI data were collected.</p> <p>This study builds on the ACT-FAST action concept taxonomy. The OSF repository for the derviation and validation of that taxonomy is linked to this project. We apply other analyses to the Sherlock fMRI dataset in that study. However, voxel selection, video annotation, and ratings are shared across the two studies. Thus see the linked repository ('fMRI Study' folder) for the corresponding code and data. For details on the code and data specific to this study, see the dictionary below.</p> <p>Data and code dictionary:</p> <p>fMRI analysis/ <em> combined_model.m - trains and tests combined (6-D) ACT-FAST decoding/prediction model. </em> combined_sbatch - slurm script for submitting combined_model.m to cluster. <em> predictive_model_discrete.m - trains and tests individual dimension decoding and prediction models. </em> predictive_model_discrete_sbatch - slurm script for submitting predictive_model_discrete.m to cluster. <em> pattern_results.R - performs NHST and visualization on results produced by combined_model.m and predictive_model_discrete.m </em> TR_regressors_annotation_332.csv - contains regressors encoding the probabilities of the 332 action classes across the 1976 TRs of the fMRI time course <em> TR_regressors_anoid_332.csv - contains numerical labels indicating the single most likely action at each TR in the fMRI timecourse. Vales correspond to the columns labels in TR_regressors_annotation_332.csv </em> TR_regressors_ratings.csv - contains regressors for the 6 ACT-FAST dimensions across the 1976 TRs of the fMRI time course.</p> <p>fMRI analysis/decoding/ <em> pls_rankmat_ratings_(1-17).csv - files containing the rank accuracy of the decoding/predictive models for individual dimensions created in predictive_model_discrete.m </em> pls_offmat_ratings_(1-17).csv -files containing the chance-level rank accuracy of the decoding/predictive models for individual dimensions created in predictive_model_discrete.m <em> pls_rankmat_multilayer_ratings_(1-17).csv - files containing the rank accuracy of the combined decoding/predictive model created in combined_model.m </em> pls_offmat_multilayer_ratings_(1-17).csv - files containing the chance-level rank accuracy of the combined decoding/predictive model created in combined_model.m</p>
OSF does not support the use of Internet Explorer. For optimal performance, please switch to another browser.
Accept
This website relies on cookies to help provide a better user experience. By clicking Accept or continuing to use the site, you agree. For more information, see our Privacy Policy and information on cookie use.
Accept
×

Start managing your projects on the OSF today.

Free and easy to use, the Open Science Framework supports the entire research lifecycle: planning, execution, reporting, archiving, and discovery.