A large-scale MEG and EEG dataset for object recognition in naturalistic scenes

Area of Science:

• Cognitive Neuroscience
• Neuroimaging
• Visual Perception

Background:

• Neuroimaging studies often use functional magnetic resonance imaging (fMRI) to understand object recognition in natural scenes.
• fMRI offers high spatial resolution but lacks the temporal dynamics crucial for analyzing rapid cognitive processes.
• Existing large-scale datasets predominantly rely on fMRI, limiting comprehensive analysis of neural mechanisms.

Purpose of the Study:

• To extend the Natural Object Dataset-fMRI (NOD-fMRI) by incorporating magnetoencephalography (MEG) and electroencephalography (EEG) data.
• To enable the examination of neural mechanisms of object recognition with both high spatial and temporal resolutions.
• To create a multimodal dataset for a more comprehensive understanding of visual object processing.

Main Methods:

• Collected fMRI, MEG, and EEG data from 30 participants viewing 57,000 naturalistic images.
• Utilized the same participants and stimuli across all neuroimaging modalities.
• Integrated existing NOD-fMRI data with newly acquired MEG and EEG recordings.

Main Results:

• The Natural Object Dataset (NOD) now comprises multimodal neuroimaging data (fMRI, MEG, EEG).
• NOD provides responses to 57,000 naturalistic images, facilitating detailed analysis of object recognition.
• The dataset enables concurrent investigation of spatial and temporal brain activity patterns.

Conclusions:

• The multimodal NOD dataset enhances the study of object recognition by combining high spatial and temporal resolution.
• NOD serves as a valuable resource for researchers investigating cognitive and neural mechanisms of visual processing.
• This integrated dataset facilitates a more holistic understanding of how the brain processes naturalistic visual information.