Personode

Personode classifies independent component analysis (ICA) maps into resting-state networks (RSNs) and derives group-specific and subject-specific regions of interest (ROIs) to improve fMRI task-related activation contrasts and resting-state functional connectivity analyses.

Key Features:

• MATLAB implementation: Packaged as a MATLAB toolbox for processing ICA maps and ROI definition.
• ICA-based network identification: Uses independent component analysis (ICA) to identify canonical resting-state networks (RSNs) while accounting for inter-individual variability.
• Semi-automated component classification: Semi-automatically classifies ICA components into RSNs for downstream ROI derivation.
• Individualized ROI generation: Produces both group-specific and subject-specific ROIs derived from RSNs.
• ROI geometries: Supports subject-specific spherical ROIs (applied to task-related activation) and subject-specific irregular ROIs (applied to resting-state analyses).
• Validation results: Demonstrated that subject-specific spherical ROIs yield higher activity contrasts than single-study and meta-analytic coordinates, and that subject-specific irregular ROIs improve functional connectivity analyses between ROIs.

Scientific Applications:

• ICA map classification: Classification of ICA components into canonical RSNs for network identification.
• Task-related activation analysis: Definition of subject-specific spherical ROIs to increase activity contrast in task fMRI studies.
• Resting-state functional connectivity: Use of subject-specific irregular ROIs to improve functional connectivity analyses in resting-state fMRI.
• Network-specific ROI delineation: Generation of group- and subject-level ROIs for more specific and accurate network analyses.

Methodology:

Independent component analysis (ICA) followed by semi-automated classification of ICA components into resting-state networks (RSNs) and derivation of group-specific and subject-specific ROIs (spherical for task data and irregular for resting-state data).