Dynamic Causal Modelling of Face Processing with fMRI and M/EEG

This dataset consists of BIDS-formatted fMRI and M/EEG data from Wakeman &amp; Henson (2015) that have been processed for DCM analysis. Multimodal data from fMRI, EEG, MEG magnetometers and MEG planar gradiometers from all 16 subjects have been processed with the analysis pipeline presented in Henson et al (2019). This dataset was prepared for demonstration of group-level estimation and inference of DCMs from fMRI and M/EEG data. The raw data can be found here: https://openneuro.org/datasets/ds000117fMRI data is provided in two formats:<i>fMRI_ProcessedData_Individual_Runs.tar.xz</i> contains processed data per run per participant, about 8.9GB total. <i>fMRI_DCMreadyData_VOI_TimeCourses.tar</i> contains VOI time courses for 3 regions - bilateral EVC, and left and right FFA, obtained after reparametrizing and concatenating processed data per run. This totals up to about 440MB.M/EEG data consists of averaged evoked sensor data for two conditions - faces and scrambled faces. Individual subjects' T1 images were used for creating head models. Forward models (leadfields) for all subjects are included in the dataset, which is provided in two formats:MEEG_DCMreadyData_with_GainMatrix.tar.xz consists of sensor data and leadfields - sufficient for inverting MEG, but not EEG data since forward models for the latter need BEM surfaces. This is compact, about 390MB.MEEG_DCMreadyData_with_GainMatrix_BEM.tar.xz consists of sensor data, leadfields and BEM surfaces - can be used for inverting either MEG or EEG data. This is considerably larger in size, about 2.5GB.All four data packages listed above have corresponding file lists for inspection, as well as SHA256 and MD5 checksums for verification of data integrity. Note that the data are highly compressed and will expand to about twice their size on decompression. Please use 7zip on Windows or `tar -xvf filename` on Linux to extract. Scripts used to produce this dataset from the raw data, as well as as scripts demonstrating group DCM analysis can be found here: https://github.com/pranaysy/MultimodalDCM.For more information or queries, please get in touch, or open an issue on the GitHub repository linked above.