Decoding sound categories based on whole-brain functional connectivity patterns

Area of Science:

• Neuroscience
• Auditory Perception
• Brain Connectivity

Background:

• Sound categorization is crucial for individuals with sensory impairments.
• Previous research used univariate or voxel-wise multivariate methods to identify brain regions sensitive to auditory categories.
• Feedback connections between brain areas are hypothesized to aid auditory object selection.

Purpose of the Study:

• To investigate if functional connectivity patterns can decode sound categories.
• To identify category-discriminative brain networks and regions involved in auditory object representation.

Main Methods:

• Constructed whole-brain functional connectivity patterns during perception of four sound categories.
• Applied multivariate pattern classification analysis for sound decoding.
• Determined discriminative networks and regions using weight maps.

Main Results:

• Achieved high accuracy in multi-category sound classification using functional connectivity patterns.
• Identified discriminative functional networks spanning across hemispheres and from primary to high-level cognitive regions.
• Superior temporal gyrus and Heschl's gyrus were significant contributors to sound category discrimination.

Conclusions:

• Functional connectivity-based multivariate classification effectively decodes auditory categories.
• Findings support a distributed representation of auditory objects, highlighting interactive properties of brain areas.
• This method offers new insights into auditory processing and representation for sensory loss patients.