Evaluation of functional MRI-based human brain parcellation: a review

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Brain Mapping

Background:

• Brain parcellations are essential for analyzing neuroimaging data, influencing study outcomes.
• Novel MRI-based parcellation methods have emerged, necessitating robust evaluation techniques.
• Evaluating functional brain parcellations derived from functional Magnetic Resonance Imaging (fMRI) data is complex due to the absence of ground truth.

Purpose of the Study:

• To review and critically assess existing methods for evaluating functional brain parcellations from fMRI data.
• To discuss the application of these evaluation methods in determining optimal parcellation granularity.
• To propose a context-dependent, teleological framework for selecting optimal brain parcellations.

Main Methods:

• Systematic review and critique of current brain parcellation evaluation methodologies.
• Analysis of how evaluation methods inform the estimation of optimal parcellation granularity.
• Development of a teleological approach for evaluating parcellations based on specific neuroimaging study applications.

Main Results:

• Current evaluation methods for functional brain parcellations are varied and have limitations.
• The concept of 'optimal' brain parcellation is application-dependent, not universal.
• A teleological approach allows for context-specific identification of suitable parcellations.

Conclusions:

• The criteria for an optimal brain parcellation must align with the intended application.
• A teleological framework, evaluating parcellations in specific contexts, is proposed for improved selection.
• Further research into advanced evaluation strategies is needed to enhance brain parcellation utility in neuroimaging.