A Comparative Study of Diffusion Fiber Reconstruction Models for Pyramidal Tract Branches.

Xinjun Suo,Wen Qin,Lining Guo,Hao Ding,Yihong Li,Dianxun Fu

doi:10.3389/fnins.2021.777377

Xinjun Suo, Wen Qin + Show 4 more

Open Access

https://doi.org/10.3389/fnins.2021.777377

Copy DOI

Abstract

Currently, comparative studies evaluating the quantification accuracy of pyramidal tracts (PT) and PT branches that were tracked based on four mainstream diffusion models are deficient. The present study aims to evaluate four mainstream models using the high-quality Human Connectome Project (HCP) dataset. Diffusion tensor imaging (DTI), diffusion spectral imaging (DSI), generalized Q-space sampling imaging (GQI), and Q-ball imaging (QBI) were used to construct the PT and PT branches in 50 healthy volunteers from the HCP. False and true PT fibers were identified based on anatomic information. One-way repeated measure analysis of variance and post hoc paired-sample t-test were performed to identify the best PT and PT branch quantification model. The number, percentage, and density of true fibers of PT obtained based on GQI and QBI were significantly larger than those based on DTI and DSI (all p < 0.0005, Bonferroni corrected), whereas false fibers yielded the opposite results (all p < 0.0005, Bonferroni corrected). More trunk branches (PTtrunk) were present in the four diffusion models compared with the upper limb (PTUlimb), lower limb (PTLlimb), and cranial (PTcranial) branches. In addition, significantly more true fibers were obtained in PTtrunk, PTUlimb, and PTLlimb based on the GQI and QBI compared with DTI and DSI (all p < 0.0005, Bonferroni corrected). Finally, GQI-based group probabilistic maps showed that the four PT branches exhibited relatively unique spatial distributions. Therefore, the GQI and QBI represent better diffusion models for the PT and PT branches. The group probabilistic maps of PT branches have been shared with the public to facilitate more precise studies on the plasticity of and the damage to the motor pathway.

Highlights

White matter fiber tracking based on the diffusion tensor imaging (DTI) model has been widely available and is often used in clinical practice and clinical neuroscience to examine the brain’s white matter (WM) microstructures (Mori and Zhang, 2006; Chen et al, 2011; Thomason and Thompson, 2011; Auriat et al, 2015; Chaudhary et al, 2015; Lucas-Jiménez et al, 2015; LopePiedrafita, 2018; Tae et al, 2018)
The DTI model has provided considerable insight into our understanding of functional connectivity based on functional magnetic resonance imaging, revealing subtle alterations in the WM microstructure between different functional networks (Cunningham et al, 2017; Li et al, 2018; Larabi et al, 2020)
Based on the high-quality Human Connectome Project (HCP) dataset, we found that generalized q-sampling imaging (GQI) and Q-ball imaging (QBI) represented better diffusion models for pyramidal tract (PT) and its branches compared with the DTI and diffusion spectrum imaging (DSI) models

Summary

Introduction

White matter fiber tracking based on the diffusion tensor imaging (DTI) model has been widely available and is often used in clinical practice and clinical neuroscience to examine the brain’s white matter (WM) microstructures (Mori and Zhang, 2006; Chen et al, 2011; Thomason and Thompson, 2011; Auriat et al, 2015; Chaudhary et al, 2015; Lucas-Jiménez et al, 2015; LopePiedrafita, 2018; Tae et al, 2018). The DTI model has provided considerable insight into our understanding of functional connectivity based on functional magnetic resonance imaging (fMRI), revealing subtle alterations in the WM microstructure between different functional networks (Cunningham et al, 2017; Li et al, 2018; Larabi et al, 2020). Many researchers have proposed a variety of improved diffusion direction quantitative models to achieve the tracking of crossed fibers and have more realistically revealed the connection of brain networks, including diffusion spectrum imaging (DSI) (Yeh and Verstynen, 2016), q-space spherical imaging (Q-ball imaging, QBI) (Descoteaux et al, 2007; Aganj et al, 2009), and generalized q-sampling imaging (GQI) (Yeh et al, 2010)

Objectives

Methods

Results

Discussion

Conclusion