A large-scale fMRI dataset for the visual processing of naturalistic scenes

Area of Science:

• Visual neuroscience
• Cognitive neuroscience
• Neuroimaging

Background:

• Understanding brain's processing of natural visual stimuli is a key goal in neuroscience.
• Large-scale functional magnetic resonance imaging (fMRI) datasets are crucial for this research.
• Existing datasets are insufficient, necessitating new resources for naturalistic stimuli.

Purpose of the Study:

• Introduce the Natural Object Dataset (NOD), a novel, large-scale fMRI dataset.
• Provide a resource for studying brain responses to naturalistic visual stimuli.
• Facilitate research on individual and stimulus variability in visual processing.

Main Methods:

• Collected fMRI data from 30 participants viewing 57,120 naturalistic images.
• Designed the dataset to balance individual differences and stimulus variations.
• Ensured comprehensive coverage of naturalistic visual stimuli.

Main Results:

• Established a large-scale fMRI dataset (NOD) with extensive naturalistic image coverage.
• The dataset captures brain responses across a diverse set of stimuli and participants.
• NOD allows for analysis of generalizability across individuals and stimuli.

Conclusions:

• The Natural Object Dataset (NOD) is a valuable resource for visual neuroscience research.
• NOD will advance understanding of how the brain processes complex natural visual information.
• This dataset will complement existing resources, driving new discoveries in visual processing.