AGFS-Tractometry: A Novel Atlas-Guided Fine-Scale Tractometry Approach for Enhanced Along-Tract Group Statistical Comparison Using Diffusion MRI Tractography

Diffusion MRI (dMRI) tractography is currently the only method for in vivo mapping of the brain’s white matter (WM) connections. Tractometry is an advanced tractography analysis technique for along-tract profiling to investigate the morphology and microstructural properties along the fiber tracts. Tractometry has become an essential tool for studying local along-tract differences between different populations (e.g., health vs disease). In this study, we propose a novel atlas-guided fine-scale tractometry method, namely AGFS-Tractometry, that leverages tract spatial information and permutation testing to enhance the along-tract statistical analysis between populations. There are two major contributions in AGFS-Tractometry. First, we create a novel atlas-guided tract profiling template that enables consistent, fine-scale, along-tract parcellation of subject-specific fiber tracts. Second, we propose a novel nonparametric permutation testing group comparison method to enable simultaneous analysis across all along-tract parcels while correcting for multiple comparisons. We perform experimental evaluations on synthetic datasets with known group differences and in vivo real data. We compare AGFS-Tractometry with two state-of-the-art tractometry methods, including Automated Fiber-tract Quantification (AFQ) and BUndle ANalytics (BUAN). Our results show that the proposed AGFS-Tractometry obtains enhanced sensitivity and specificity in detecting local WM differences. In the real data analysis experiments, AGFS-Tractometry can identify more regions with significant differences, which are anatomically consistent with the existing literature. Overall, these demonstrate the ability of AGFS-Tractometry to detect subtle or spatially localized WM group-level differences. The created tract profiling template and related code are available at: https://github.com/ZhengRuixi/AGFS-Tractometry.git.

💡 Research Summary

AGFS‑Tractometry introduces a novel, atlas‑guided fine‑scale tractometry framework that substantially improves the spatial resolution and statistical power of along‑tract analyses in diffusion MRI (dMRI) tractography. The method builds on the ORG‑atlas, an anatomically curated white‑matter fiber clustering atlas derived from dense tractography of 100 subjects. Each fiber cluster in the atlas is first represented by a centerline obtained through equidistant resampling of all streamlines within the cluster. The centerline is then divided into n equally spaced points (e.g., n = 100), and every streamline point is assigned to its nearest centerline point, yielding n parcels per cluster. Consequently, a single anatomical tract composed of m clusters is partitioned into m × n parcels (e.g., 15 × 100 = 1500 parcels), providing a fine‑scale along‑tract parcellation that captures both longitudinal and transverse anatomical variation.

A key innovation lies in the construction of a parcel‑wise spatial neighborhood matrix. Two parcels are considered neighbors if they belong to the same cluster and are adjacent along the centerline, or if they belong to different clusters but their centerline points are closer than the sum of their radii (the average distance of points in a parcel to its centerline). This matrix encodes intra‑ and inter‑cluster spatial continuity, preserving structural integrity while allowing for very fine subdivisions.

Statistical comparison proceeds in three steps. First, an uncorrected parcel‑wise test (e.g., Student’s t‑test or linear mixed‑effects model) yields a p‑value for each parcel. Second, parcels with p < 0.05 are linked according to the neighborhood matrix, forming connected components (communities). Third, a non‑parametric permutation test evaluates each community’s summary statistic (size, mean test statistic) against a null distribution generated by randomly permuting group labels across subjects (5 000–10 000 permutations). Communities that exceed the 95th percentile of the null distribution are declared significant, providing a cluster‑thresholding correction that leverages spatial contiguity rather than treating parcels independently.

The authors validate AGFS‑Tractometry on three fronts. In synthetic data with known ground‑truth differences, the method accurately recovers the altered parcels while maintaining low false‑positive rates. In a large population dataset examining sex differences, AGFS‑Tractometry identifies more significant parcels than AFQ and BUAN, notably resolving the corticospinal tract (CST) into sub‑divisions (leg, trunk, hand, face) and detecting sex‑related microstructural variations within each sub‑division. In a clinical application comparing autism spectrum disorder (ASD) patients to controls, the approach uncovers additional localized white‑matter differences beyond those reported in prior literature, aligning with known neuroanatomical alterations in ASD.

Overall, AGFS‑Tractometry offers two major contributions: (1) an atlas‑guided fine‑scale tract profiling template that ensures consistent, high‑resolution parcellation across subjects, and (2) a spatially informed, permutation‑based group comparison that simultaneously tests all parcels while controlling for multiple comparisons through cluster‑thresholding. The method demonstrates superior sensitivity and specificity relative to state‑of‑the‑art tractometry pipelines (AFQ, BUAN). All code and the tract profiling template are publicly released (https://github.com/ZhengRuixi/AGFS‑Tractometry.git), facilitating reproducibility and encouraging adoption in future neuroimaging studies of development, disease, and individual differences.