Simultaneous EEG-fMRI dataset
Description
Here we present an open-access multimodal neuroimaging dataset consisting of simultaneously and independently collected electroencephalographic (EEG) and Magnetic Resonance Imaging (MRI) data from 20 healthy young male adults (mean age=26 years; SD=3.8 years). The dataset follows the BIDS standard specification and is organized into two parts: 1) EEG data collected outside the Magnetic Resonance (MR) environment, inside the MR scanner without acquiring images and during the simultaneous acquisition of functional MRI, which also includes the gradient artifact-corrected version of the EEG data (_gac suffix) and 2) the fMRI data collected during the simultaneous EEG-fMRI acquisitions, the fMRI data collected without the EEG cap, and the structural MRI data acquired with and without the EEG cap. EEG data were recorded using an MR-compatible EEG recording system (GES 400 MR, Electrical Geodesics Inc.) equipped with a 32-channel sponge-based EEG cap. 2 minutes of eyes-closed (EC) resting-state EEG data were recorded for the outside and inside scanner conditions, while 10 minutes were recorded for the EEG-fMRI condition. 2 minutes of eyes-open (EO) resting-state EEG were recorded for the 3 conditions. Subjects additionally performed an eyes open-closed (EO-EC) block-design task outside the scanner (2 minutes) and during simultaneous EEG-fMRI (4 minutes). The outside EEG data provides a baseline condition of the signal without the presence of MR-related artifacts. The inside EEG data records the contribution of the ballistocardiographic (BCG) artifact without the presence of the gradient artifact, representing an ideal condition to test and validate BCG artifact correction methods. The EEG-fMRI condition records the contribution of both the gradient and the BCG artifact. Brain images were obtained using a 3T GE MR750-Discovery MR scanner equipped with a 32-channel head coil. Whole-brain functional images were obtained using a GRE-EPI T2* weighted sequence (TR = 2000 ms, TE = 40 ms, 35 interleaved axial slices with 4 mm isometric voxels). Structural images were acquired with a SPGR sequence (TR = 8.1 ms, TE = 3.2 ms, flip angle = 12°, 176 sagittal slices with 1 mm isometric voxels). Anatomical data were always collected after their corresponding functional acquisitions. This represents one of the largest open-access EEG-fMRI datasets available, also including independently collected EEG and structural/functional MRI data. This dataset will allow other researchers to 1) characterize the impact of the gradient and the BCG artifact on low-density EEG data, 2) use these data to assess the effectiveness of novel artifact removal approaches to reduce the artifacts contribution and preserve EEG signal properties, 3) perform hardware-setup comparison studies, 4) address the quality of structural and functional MRI data collected along with the EEG cap, and 5) implement and validate multimodal integrative analysis on the EEG-fMRI data acquired simultaneously.
Files
Steps to reproduce
This dataset comprises multimodal neuroimaging data from 20 healthy male subjects (mean/std age = 26/3.8 years). Subjects were administered the MINI International Neuropsychiatric Interview and only subjects who did not have diagnosis of neurological/psychiatric disease or history of substance abuse were invited to participate. This research project was conducted in accordance with the principles of the Declaration of Helsinki and approved by the Bioethics Committee of our institution. Both EEG and MRI data were acquired in a single session. EEG data were recorded using a GES 400 MR-compatible system equipped with a 32-channel Geodesic Sensor Net EEG cap (Electrical Geodesics Inc., OR, USA). A silk mesh was placed over the cap to reduce electrode movement and improve EEG data quality. Electrode impedances were kept below 50 k-ohms. Sampling rate was 1000 Hz and Cz was used as reference electrode. Electrocardiogram was recorded using MR-compatible patch electrodes. Data were recorded using Net Station software v. 5.3. EEG data were first collected outside the scanner. Eyes closed (EC) EEG data were recorded for 2 minutes, followed by 2 minutes of eyes open (EO) and 2 minutes of alternating between eyes open and eyes closed states (EO-EC task). The subject was taken into the MR scanner room. The EEG amplifier was placed next to the scanner bore, EEG leads were examined in search of loops and oriented parallel to the B0 magnetic field, and sandbags and tape were placed over the EEG leads. Soft pads were used to minimize the subject’s head movement. Lights and ventilation systems were turned off during the entire session, while the helium pump remained on during all recording sessions. Once the subject felt comfortable inside the scanner, we recorded 2 minutes of EC and 2 minutes of EO EEG without image acquisition. We then began the simultaneous EEG-fMRI protocol. Brain images were obtained with a Discovery-MR750 3.0 T MR scanner (General Electric, WI, USA), equipped with a 32-channel array head coil. BOLD contrast functional images were acquired using a Gradient Recalled Echoplanar Imaging T2* sequence, and anatomical acquisitions were collected using a Spoiled Gradient Recalled sequence. We recorded an EC resting-state condition for 10 minutes, followed by a brief 2-minute EO resting-state condition, 4 minutes of the EO-EC task, and the structural scan acquisition. After removing the EEG cap, we repeated the acquisition of the 10-minute EC resting state and the anatomical scan. Raw data was adapted according to the BIDS specification standard. Raw EEG files were exported to the EEGLAB data structure, where channel locations and labels were specified. BIDS sidecar files were created using the data2bids function included in the Fieldtrip software package. Raw MRI DICOM files were converted to nifti file format and transformed to BIDS using the heuristic DICOM converter. The json files contain relevant metadata information of the scanning sessions.