Video-evoked fMRI BOLD responses are highly consistent across different data acquisition sites

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Neuroimaging

Background:

• Naturalistic imaging paradigms using video stimuli are common in cognitive neuroscience.
• Videos elicit synchronized brain responses within and across subjects.
• Reproducibility of these brain responses across different data acquisition sites remains unquantified.

Purpose of the Study:

• To quantify the consistency of video-evoked brain responses across independent functional magnetic resonance imaging (fMRI) sites.
• To compare brain response reproducibility between matched and unmatched fMRI datasets.
• To assess the generalizability of video fMRI studies.

Main Methods:

• Compared brain responses from participants viewing identical videos at two different fMRI sites (Indiana University and Caltech).
• Utilized two matched datasets with identical scanner, acquisition, and preprocessing details.
• Included a third unmatched dataset with varying details to assess robustness.

Main Results:

• Video-evoked brain responses demonstrated high consistency across different fMRI sites for both matched and unmatched datasets.
• Site differences were more pronounced in unmatched datasets compared to matched datasets.
• Residual differences may stem from participant-level variability.

Conclusions:

• Video fMRI studies exhibit high cross-site reproducibility, indicating robustness.
• Findings support the development and generalization of video fMRI for studying individual differences in healthy and clinical populations.
• High consistency across sites facilitates multi-site collaborations and broader application of video fMRI research.