Evaluating the Reliability of Human Brain White Matter Tractometry

Abstract

The validity of research results depends on the reliability of analysis methods. In recent years, there have been concerns about the validity of research that uses diffusion-weighted MRI (dMRI) to understand human brain white matter connections in vivo, in part based on the reliability of analysis methods used in this field. We defined and assessed three dimensions of reliability in dMRI-based tractometry, an analysis technique that assesses the physical properties of white matter pathways: (1) reproducibility, (2) test-retest reliability, and (3) robustness. To facilitate reproducibility, we provide software that automates tractometry (https://yeatmanlab.github.io/pyAFQ). In measurements from the Human Connectome Project, as well as clinical-grade measurements, we find that tractometry has high test-retest reliability that is comparable to most standardized clinical assessment tools. We find that tractometry is also robust: showing high reliability with different choices of analysis algorithms. Taken together, our results suggest that tractometry is a reliable approach to analysis of white matter connections. The overall approach taken here both demonstrates the specific trustworthiness of tractometry analysis and outlines what researchers can do to establish the reliability of computational analysis pipelines in neuroimaging.

White matter and literacy: a dynamic system in flux

“…It should be noted that the diffusion MRI data for the HBN, ABCD, and HCP-YA samples were analyzed using pyAFQ and the PLING, PING, and ABCD samples were analyzed using AtlasTrack. Although tractometry was performed using different computational pipelines, past studies have shown that these analyses are robust to the details of the methodology (43). Code to reproduce all the results in the notebook is available: https://github.com/earoy/longitudinal_wm.…”

“…Once the diffusion imaging data were preprocessed, pyAFQ was used to calculate tractometry properties (43). Briefly, constrained spherical deconvolution (77), implemented in DIPY (78) was used to estimate fiber orientation distributions in every voxel, and probabilistic tractography was used to propagate streamlines throughout the white matter.…”

Specific and non-linear effects of glaucoma on optic radiation tissue properties

“…AFQ 21 is an analysis pipeline that automatically delineates major white matter pathways and quantifies the properties of white matter tissue along the length of these major pathways. We used a Python-based open-source implementation of this pipeline (pyAFQ; https://github.com/yeatmanlab/pyAFQ) 22 to extract the tissue properties of OR sub-bundles and two control bundles – the corticospinal tract (CST) and the uncinate fasciculus (UNC) – that are not a part of the visual system. In AFQ, white matter pathways are identified from a candidate set of streamlines based on anatomical landmarks: inclusion, exclusion, and endpoint regions of interest (ROIs) that are based on the known trajectory of the pathway (e.g., OR).…”

“…In the present work, we used dMRI data from the UKBB to compare participants with glaucoma to a statistically matched sample of participants that do not have glaucoma. To quantify tissue properties, we used Automated Fiber Quantification 21 , an automated method, implemented in open-source software that we have developed 22 , that delineates the OR in each individual. Tissue properties of the OR were quantified using the diffusional kurtosis imaging model (DKI) 23 24 .…”

A comparison of diffusion tensor imaging tractography and constrained spherical deconvolution with automatic segmentation in traumatic brain injury

“…TractSeg is robust in eliminating false positive fibres, which would otherwise be a problem with probabilistic CSD-based tractography ( Maier-Hein et al, 2017 ). Other semi-automated tools for tract segmentation exist, including AFQ/pyAFQ ( Kruper et al, 2021 , Yeatman et al, 2012 ), and WMA segmentation ( Bullock et al, 2019 ).…”