Sparse imaging and continuous event-related fMRI in the visual domain: a systematic comparison

Area of Science:

• Neuroimaging
• Cognitive Neuroscience
• Biomedical Engineering

Background:

• Continuous functional magnetic resonance imaging (fMRI) acquisition can be limited by gradient switching noise.
• Sparse fMRI acquires single images with a delay to capture the peak hemodynamic response (HDR).

Purpose of the Study:

• To evaluate the validity and sensitivity of single-trial sparse fMRI in the visual domain.
• To compare sparse fMRI with continuous event-related fMRI.

Main Methods:

• Thirteen subjects underwent two scanning sessions, one with continuous and one with sparse fMRI, viewing flashlight stimuli.
• Activation patterns (spatial extent, overlap, intensity) were compared between acquisition methods.
• HDR peak latency variability was analyzed for continuous fMRI to inform sparse fMRI timing.
• Sparse fMRI sensitivity was assessed with varying numbers of averaged trials.

Main Results:

• Sparse fMRI successfully detected core brain activity identified by continuous fMRI.
• Signal change intensity was comparable between sparse and continuous fMRI.
• HDR peak latency showed session stability but intersubject variability.
• Increasing averaged trials in sparse fMRI improved statistical power and extended activation detection.

Conclusions:

• Sparse fMRI is a valid neuroimaging technique, comparable to continuous fMRI for detecting visual activation.
• Continuous fMRI provides more comprehensive data, but sparse fMRI allows for artifact-free stimulus presentation and biosignal recording.
• Sparse fMRI's utility is influenced by HDR peak latency variability and benefits from trial averaging.