Sparse and continuous sampling approaches to fMRI of overt vocalization tasks

Area of Science:

• Neuroimaging
• Auditory Neuroscience
• Speech Science

Background:

• Sparse temporal sampling is dominant in fMRI for auditory/verbal tasks, enabling stimulus presentation during quiet scanner periods.
• No direct comparison exists for sparse vs. continuous acquisition in overt auditory-verbal tasks like singing.
• Continuous acquisition may offer advantages for certain studies, but its utility for overt responses is less explored.

Purpose of the Study:

• To compare sparse and continuous fMRI acquisition for overt singing tasks.
• To evaluate if sparse sampling reduces movement artifacts and improves network activation.
• To assess the impact of acquisition type on image quality and downstream analysis.

Main Methods:

• Fifteen healthy adults performed an overt singing task under both sparse and continuous fMRI scanning conditions.
• Data were analyzed for signal intensity variations, slice-specific artifacts, and movement parameters.
• Functional activation differences between sparse and continuous data were statistically compared.

Main Results:

• Sparse acquisition showed significant slice-intensity variations and lower activation in later-acquired slices, necessitating slice removal.
• Edge artifacts were present in both, with more ventricular space artifacts in continuous data.
• No significant differences in functional activation or motion correction parameters were found between the two methods.

Conclusions:

• Sparse fMRI sampling can introduce significant image artifacts impacting analyses, requiring careful data inspection.
• Continuous fMRI acquisition is a feasible and potentially more versatile method for overt auditory-verbal studies.
• The technical demands of sparse sequences should not be underestimated, while continuous imaging may be more broadly applicable.