Shape analysis of the human association pathways

• Published in: NeuroImage (Orlando, Fla.) Vol. 223; p. 117329
• Main Author: Yeh, Fang-Cheng
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.12.2020; Elsevier Limited; Elsevier
• Subjects: Adult; Anisotropy; Automatic fiber tracking; Brain; Brain - anatomy & histology; Brain - diagnostic imaging; Brain Mapping - methods; Cingulum; Computer vision; Consortia; Diffusion MRI; Diffusion Tensor Imaging; Dominance; Female; Humans; Image processing; Image Processing, Computer-Assisted - methods; Innervation; Investigations; Magnetic resonance imaging; Male; Morphology; Neural Pathways - anatomy & histology; Neural Pathways - diagnostic imaging; Shape analysis; Shape descriptor; Structure-function relationships; Tractography; Variation; White Matter - anatomy & histology; White Matter - diagnostic imaging; Young Adult; Diffusion MRI; Automatic fiber tracking; Shape analysis; Shape descriptor; Tractography
• DOI: 10.1016/j.neuroimage.2020.117329
• ISSN: 1053-8119; 1095-9572

Shape analysis has been widely used in digital image processing and computer vision, but they have not been utilized to compare the structural characteristics of the human association pathways. Here we used shape analysis to derive length, area, volume, and shape metrics from diffusion MRI tractography and utilized them to study the morphology of human association pathways. The reliability analysis showed that shape descriptors achieved moderate to good test-retest reliability. Further analysis on association pathways showed left dominance in the arcuate fasciculus, cingulum, uncinate fasciculus, frontal aslant tract, and right dominance in the inferior fronto-occipital fasciculus and inferior longitudinal fasciculus. The superior longitudinal fasciculus has a mixed lateralization profile with different metrics showing either left or right dominance. The analysis of between-subject variations shows that the overall layout of the association pathways does not variate a lot across subjects, as shown by low between-subject variation in length, span, diameter, and radius. In contrast, the area of the pathway innervation region has a considerable between-subject variation. A follow-up analysis is warranted to thoroughly investigate the nature of population variations and their structure-function correlation. [Display omitted]