Related Experiment Videos
Separating processes within a trial in event-related functional MRI II. Analysis.
J M Ollinger1, M Corbetta, G L Shulman
1Department of Radiology, Washington University, St. Louis, Missouri, 63110, USA. jmo@npg.wustl.edu
Neuroimage
|January 3, 2001
Summary
Accurate estimation of blood oxygenation-level-dependent (BOLD) responses in functional MRI requires careful experimental design. Optimizing intertrial intervals and using appropriate statistical models, like the gamma function, improve signal interpretation.
Area of Science:
- Neuroimaging
- Cognitive Neuroscience
- Functional Magnetic Resonance Imaging (fMRI)
Background:
- Cognitive processes unfold over time scales impacting the blood oxygenation-level-dependent (BOLD) response in event-related fMRI.
- Interpreting BOLD time courses requires careful consideration of noise, experimental design, and statistical modeling.
Purpose of the Study:
- To investigate optimal strategies for estimating BOLD response shapes in event-related fMRI.
- To identify key factors influencing the accuracy of BOLD time course estimation.
Main Methods:
- Simulations were used to evaluate the impact of intertrial interval (ITI) timing and trial types (full vs. partial) on BOLD signal estimation.
- Statistical mapping approaches, including F statistics and t statistics derived from cross-correlation with a hemodynamic response model, were compared.
- The robustness of different hemodynamic response models, specifically a single gamma function, was assessed.
Main Results:
- Optimal BOLD response estimation is achieved with short, exponentially distributed ITIs (at least three values) and a significant proportion of partial trials (20-40%).
- The accuracy of time course estimation is relatively robust to variations in the percentage of partial trials.
- A single gamma function proved to be the most reliable model for the hemodynamic response when using cross-correlation methods.
- High-pass filtering is beneficial for rapid event-related fMRI paradigms when an appropriate hemodynamic model is employed.
Conclusions:
- Careful selection of experimental design parameters, particularly ITI timing and trial structure, is crucial for accurate BOLD response estimation in event-related fMRI.
- The gamma function provides a robust model for the hemodynamic response, enhancing statistical mapping accuracy.
- Understanding these factors improves the interpretation of fMRI data in cognitive neuroscience research.