Laminar EEG-fMRI of Visual Features (Clausner et al.)

Simultaneous EEG and high-resolution fMRI of the human visual cortex, recorded at the Donders Institute in Nijmegen by Tommy Clausner, José Marques, René Scheeringa and Mathilde Bonnefond. Healthy adults watched gratings tilted left or right and reported rare wavy oddball patterns while EEG and submillimetre BOLD images were recorded in alternation. The authors related trial-by-trial alpha and gamma power to the BOLD signal in superficial, middle and deep layers of V1, V2 and V3. They report that gamma power goes with orientation-specific BOLD increases, while alpha splits into a general lower-frequency component and a feature-specific upper-frequency one. The collection sits in the Radboud Data Repository behind a data use agreement for bona fide researchers.

Composition

52 right-handed adults aged 18 to 35 (mean 24, SD 4), 34 of them female. None had a history of neurological or psychiatric illness or neurosurgery, and all could see without glasses. No participant was excluded. According to the paper, the deposit holds the raw EEG, fMRI and eye tracking data together with all derivatives. The repository lists a total size of about 1.6 TB.

Acquisition

One Siemens MAGNETOM Prismafit 3 T with a 64-channel head and neck coil. A 0.8 mm MPRAGE was acquired during a practice block. Functional images came from a 3D gradient-echo EPI with CAIPI acceleration over a coronal slab of occipital and parietal cortex, with 0.9 mm voxels, a volume TR of 3.3 s and a TE of 34 ms. In the four 14-minute task blocks a 3 s gap followed every three volumes so that EEG could be recorded free of gradient artefacts. Three 7-minute population receptive field mapping blocks used the same sequence without gaps. A short series with inverted flip angle was taken for distortion estimation. EEG came from a 64-channel MR-compatible Brain Products system at 5 kHz; electrode positions were measured by photogrammetry, and only the final sensor locations are shared. Eye movements were tracked with an EyeLink 1000 Plus.

Known limitations

• The fMRI slab covers only the occipital and parietal lobes.
• The eye tracking data were judged too poor to analyse.
• The sample is small, young and from one site.
• Access requires an account and acceptance of a scientific-use agreement; the data may not be passed on.