A critical review of connectome validation studies

Abstract

Diffusion MRI tractography is the most widely used macroscale method for mapping connectomes in vivo. However, tractography is prone to various errors and biases, and thus tractography-derived connectomes require careful validation. Here, we critically review studies that have developed or utilized phantoms and tracer maps to validate tractography-derived connectomes, either quantitatively or qualitatively. We identify key factors impacting connectome reconstruction accuracy, including streamline seeding, propagation and filtering methods, and consider the strengths and limitations of state-of-the-art connectome phantoms and associated validation studies. These studies demonstrate the inherent limitations of current fiber orientation models and tractography algorithms and their impact on connectome reconstruction accuracy. Reconstructing connectomes with both high sensitivity and high specificity is challenging, given that some tractography methods can generate an abundance of spurious connections, while others can overlook genuine fiber bundles.We argue that streamline filtering can minimize spurious connections and potentially improve the biological plausibility of connectomes derived from tractography. We find that algorithmic choices such as the tractography seeding methodology, angular threshold, and streamline propagation method can substantially impact connectome reconstruction accuracy. Hence, careful application of tractography is necessary to reconstruct accurate connectomes. Improvements in diffusion MRI acquisition techniques will not necessarily overcome current tractography limitations without accompanying modeling and algorithmic advances. | INTRODUCTIONDiffusion-weighted MRI (dMRI) can be used to non-invasively infer microstructural properties of the brain. Tractography is a reconstruction method based on dMRI to compute the trajectories of white matter (WM) fiber bundles. The primary objective of tractography is to reconstruct non-invasively WM fiber bundles from dMRI. More recently, tractography has been used to map the human structural connectome 1-3 i.e. a comprehensive network description of WM fiber bundles connecting pairs of gray matter (GM) regions (also referred to as structural connectivity).Structural connectomes model the physical links between brain regions, whereas functional connectomes represent the network of functionally connected regions. These connectomes are a static representation of the brain that could change over time. 4,5 When referring to connectomes in this review, we specifically mean structural connectomes mapped with tractography.Generally, connectome mapping using tractography is a five-step process involving (1) dMRI acquisition, (2) segmenting the cortex into regions of interest (ROIs)-cortical parcellation, (3) diffusion/fiber orientation estimation, (4) streamline generation using tractography, and

Spatially heterogeneous structure-function coupling in haemodynamic and electromagnetic brain networks

“…The present results should be considered in light of two important methodological limitations. First, structural connectivity was estimated from diffusion MRI, a technique that is affected by systematic false positives and false negatives [21, 30, 35, 36, 57, 58, 69, 81, 99, 111, 118, 122, 132, 143, 160, 164]. We attempted to mitigate this issue by deriving a group consensus structural connectivity network that identifies consistent connections across many participants, but improvements in imaging technology and computational tractometry are still necessary.…”

Cortical Cross-Frequency Coupling Is Affected by in utero Exposure to Antidepressant Medication

“…However, the currently available animal literature suggests that cellular level effects of SRIs are found in the intracortical microstructure ( Persico et al, 2001 ; Xu et al, 2004 ; Liao and Lee, 2011 ; Simpson et al, 2011 ; Suri et al, 2015 ), which is an order of magnitude smaller detail than what can be genuinely studied with MR imaging. In addition, the cellular level correlates of the MR results are too poorly known to allow their use as a genuinely translational biomarker ( Jones et al, 2013 ; Sarwar et al, 2021 ). Instead, newborn brain function can be studied directly by measuring neuronal population activity with scalp electroencephalography (EEG) or indirectly by measuring blood-oxygen-level-dependent (BOLD) signal fluctuations with an fMRI method.…”

What Models and Tools can Contribute to a Better Understanding of Brain Activity?

“…Approaches that characterize properties of interactions form the basis (so called "functional" and "effective" connectivity) of a data-driven construction of functional brain networks (Eguıluz et al, 2005;Bullmore and Sporns, 2009;Bullmore and Sporns, 2012;Stam, 2014;Bassett and Sporns, 2017;Lynn and Bassett, 2019). These approaches are backed up with a variety of imaging techniques (Sporns, 2011;Fornito et al, 2015;Rockland, 2015;Fornito et al, 2019;Sotiropoulos and Zalesky, 2019;Sarwar et al, 2021;Yeh et al, 2021;Gosak et al, 2022) that aim at characterizing the so called "structural" connectivity, which is often considered ground truth and underlying constraint on "functional"/"effective" connectivity. Structural information is mostly derived from magnetic-imaging-based techniques such as diffusion tensor imaging (DTI).…”