Test-retest reliability of fMRI activation generated by different saccade tasks

Area of Science:

• Neuroscience
• Cognitive Science
• Neuroimaging

Background:

• Reproducibility is crucial for validating neuroimaging findings.
• Saccade tasks are commonly used to study eye movements and associated brain activity.
• Assessing variability across subjects, tasks, and sites is essential for robust research.

Purpose of the Study:

• To evaluate the reproducibility of brain activation and eye movement patterns during saccade tasks.
• To compare reproducibility across different levels of target predictability (predictable, position predictable, time predictable, prosaccade).
• To assess the influence of subject, task, and magnetic resonance (MR) system variations on reproducibility.

Main Methods:

• Forty-five healthy adults participated at two different sites.
• Participants performed four saccade tasks with varying target predictability.
• Reproducibility was assessed using eye-movement analysis and activation map comparisons (Jaccard index, coefficient of determination).

Main Results:

• Eye movement latencies increased with decreasing target predictability.
• Activation overlap was good to fair within-subjects but poor for time-predictable saccades between subjects.
• Activation consistency ranged from 50-85% across different comparison groups, with prosaccade tasks showing higher consistency.

Conclusions:

• Predictive saccade tasks demonstrate reliable brain activation and eye movement patterns.
• These findings support the use of predictive saccade paradigms for reproducible neuroscience research.
• The study highlights the importance of considering subject and site variations in neuroimaging studies.