[Paper Review] A Review on MR Based Human Brain Parcellation Methods

Pantea Moghimi, Anh The Dang+3 more|arXiv (Cornell University)|Jul 7, 2021

Advanced MRI Techniques and Applications199 references4 citations

TL;DR

This paper provides a comprehensive review of MRI-based human brain parcellation methods, categorizing them into anatomical (T1-weighted MRI), functional (fMRI), and structural (diffusion-weighted imaging) parcellations. It introduces a multi-level taxonomy to organize the literature, compares methodological strengths and weaknesses, and identifies key challenges in developing robust, standardized brain parcellations across modalities.

ABSTRACT

Brain parcellations play a ubiquitous role in the analysis of magnetic resonance imaging (MRI) datasets. Over 100 years of research has been conducted in pursuit of an ideal brain parcellation. Different methods have been developed and studied for constructing brain parcellations using different imaging modalities. More recently, several data-driven parcellation methods have been adopted from data mining, machine learning, and statistics communities. With contributions from different scientific fields, there is a rich body of literature that needs to be examined to appreciate the breadth of existing research and the gaps that need to be investigated. In this work, we review the large body of in vivo brain parcellation research spanning different neuroimaging modalities and methods. A key contribution of this work is a semantic organization of this large body of work into different taxonomies, making it easy to understand the breadth and depth of the brain parcellation literature. Specifically, we categorized the existing parcellations into three groups: Anatomical parcellations, functional parcellations, and structural parcellations which are constructed using T1-weighted MRI, functional MRI (fMRI), and diffusion-weighted imaging (DWI) datasets, respectively. We provide a multi-level taxonomy of different methods studied in each of these categories, compare their relative strengths and weaknesses, and highlight the challenges currently faced for the development of brain parcellations.

Motivation & Objective

• To synthesize and organize the extensive body of research on in vivo brain parcellation using MRI modalities.
• To categorize existing brain parcellation methods into anatomical, functional, and structural types based on imaging modality.
• To develop a multi-level taxonomy that clarifies the relationships and distinctions among data-driven, machine learning, and statistical parcellation techniques.
• To compare the relative strengths and limitations of different parcellation methods across neuroimaging modalities.
• To identify open challenges and research gaps in the development of standardized, reliable, and interpretable brain parcellations.

Proposed method

• The authors conducted a systematic review of over a century of brain parcellation research, focusing on in vivo MRI-based methods.
• They classified parcellation methods into three main categories: anatomical (T1-weighted MRI), functional (fMRI), and structural (diffusion-weighted imaging).
• A multi-level taxonomy was developed to organize methods within each category based on underlying principles, such as clustering, atlas-based, or machine learning approaches.
• The review integrates contributions from neuroscience, data mining, machine learning, and statistics to provide a cross-disciplinary perspective.
• Strengths and weaknesses of each method category were evaluated based on reproducibility, biological plausibility, and technical robustness.
• The paper highlights methodological inconsistencies and the lack of standardized benchmarks across studies.

Experimental results

Research questions

• RQ1 How have brain parcellation methods evolved across different MRI modalities over the past century?
• RQ2 What are the key methodological differences between anatomical, functional, and structural parcellation approaches?
• RQ3 How do data-driven techniques from machine learning and statistics compare in performance and interpretability to traditional atlas-based methods?
• RQ4 What are the major challenges hindering the development of a universally accepted brain parcellation framework?
• RQ5 What gaps remain in the current literature regarding reproducibility, validation, and standardization of parcellation methods?

Key findings

• The review identifies over 100 years of research on brain parcellation, with significant methodological diversity across anatomical, functional, and structural modalities.
• Anatomical parcellations based on T1-weighted MRI are widely used but often lack functional or connectivity-based validation.
• Functional parcellations derived from fMRI show strong correlation with functional networks but are sensitive to preprocessing and hemodynamic variability.
• Structural parcellations using diffusion-weighted imaging provide connectivity-based parcellations but require complex tractography and are computationally intensive.
• Despite advances, no consensus exists on the optimal parcellation scheme, and methodological heterogeneity limits cross-study comparability.
• The authors highlight a critical need for standardized benchmarks, validation frameworks, and open-access repositories to improve reproducibility and integration across studies.