Impact of thresholding on the consistency and sensitivity of diffusion MRI-based brain networks in patients with cerebral small vessel disease.

INTRODUCTION: Thresholding of low-weight connections of diffusion MRI-based brain networks has been proposed to remove false-positive connections. It has been previously established that this yields more reproducible scan-rescan network architecture in healthy subjects. In patients with brain disease, network measures are applied to assess inter-individual variation and changes over time. Our aim was to investigate whether thresholding also achieves improved consistency in network architecture in patients, while maintaining sensitivity to disease effects for these applications. METHODS: We applied fixed-density and absolute thresholding on brain networks in patients with cerebral small vessel disease (SVD, n = 86; ≈24 months follow-up), as a clinically relevant exemplar condition. In parallel, we applied the same methods in healthy young subjects (n = 44; scan-rescan interval ≈4 months) as a frame of reference. Consistency of network architecture was assessed with dice similarity of edges and intraclass correlation coefficient (ICC) of edge-weights and hub-scores. Sensitivity to disease effects in patients was assessed by evaluating interindividual variation, changes over time, and differences between those with high and low white matter hyperintensity burden, using correlation analyses and mixed ANOVA. RESULTS: Compared to unthresholded networks, both thresholding methods generated more consistent architecture over time in patients (unthresholded: dice = .70; ICC: .70-.78; thresholded: dice = .77; ICC: .73-.83). However, absolute thresholding created fragmented nodes. Similar observations were made in the reference group. Regarding sensitivity to disease effects in patients, fixed-density thresholds that were optimal in terms of consistency (densities: .10-.30) preserved interindividual variation in global efficiency and node strength as well as the sensitivity to detect effects of time and group. Absolute thresholding produced larger fluctuations of interindividual variation. CONCLUSIONS: Our results indicate that thresholding of low-weight connections, particularly when using fixed-density thresholding, results in more consistent network architecture in patients with longer rescan intervals, while preserving sensitivity to disease effects.

Improved sensitivity and precision in multicentre diffusion MRI network analysis using thresholding and harmonization.

PURPOSE: To investigate if network thresholding and raw data harmonization improve consistency of diffusion MRI (dMRI)-based brain networks while also increasing precision and sensitivity to detect disease effects in multicentre datasets. METHODS: Brain networks were reconstructed from dMRI of five samples with cerebral small vessel disease (SVD; 629 patients, 166 controls), as a clinically relevant exemplar condition for studies on network integrity. We evaluated consistency of network architecture in age-matched controls, by calculating cross-site differences in connection probability and fractional anisotropy (FA). Subsequently we evaluated precision and sensitivity to disease effects by identifying connections with low FA in sporadic SVD patients relative to controls, using more severely affected patients with a pure form of genetically defined SVD as reference. RESULTS: In controls, thresholding and harmonization improved consistency of network architecture, minimizing cross-site differences in connection probability and FA. In patients relative to controls, thresholding improved precision to detect disrupted connections by removing false positive connections (precision, before: 0.09-0.19; after: 0.38-0.70). Before harmonization, sensitivity was low within individual sites, with few connections surviving multiple testing correction (k = 0-25 connections). Harmonization and pooling improved sensitivity (k = 38), while also achieving higher precision when combined with thresholding (0.97). CONCLUSION: We demonstrated that network consistency, precision and sensitivity to detect disease effects in SVD are improved by thresholding and harmonization. We recommend introducing these techniques to leverage large existing multicentre datasets to better understand the impact of disease on brain networks.

Diffusion MRI harmonization enables joint-analysis of multicentre data of patients with cerebral small vessel disease.

OBJECTIVES: Acquisition-related differences in diffusion magnetic resonance imaging (dMRI) hamper pooling of multicentre data to achieve large sample sizes. A promising solution is to harmonize the raw diffusion signal using rotation invariant spherical harmonic (RISH) features, but this has not been tested in elderly subjects. Here we aimed to establish if RISH harmonization effectively removes acquisition-related differences in multicentre dMRI of elderly subjects with cerebral small vessel disease (SVD), while preserving sensitivity to disease effects. METHODS: Five cohorts of patients with SVD (N = 397) and elderly controls (N = 175) with 3 Tesla MRI on different systems were included. First, to establish effectiveness of harmonization, the RISH method was trained with data of 13 to 15 age and sex-matched controls from each site. Fractional anisotropy (FA) and mean diffusivity (MD) were compared in matched controls between sites using tract-based spatial statistics (TBSS) and voxel-wise analysis, before and after harmonization. Second, to assess sensitivity to disease effects, we examined whether the contrast (effect sizes of FA, MD and peak width of skeletonized MD - PSMD) between patients and controls within each site remained unaffected by harmonization. Finally, we evaluated the association between white matter hyperintensity (WMH) burden, FA, MD and PSMD using linear regression analyses both within individual cohorts as well as with pooled scans from multiple sites, before and after harmonization. RESULTS: Before harmonization, significant differences in FA and MD were observed between matched controls of different sites (p < 0.05). After harmonization these site-differences were removed. Within each site, RISH harmonization did not alter the effect sizes of FA, MD and PSMD between patients and controls (relative change in Cohen's d = 4 %) nor the strength of association with WMH volume (relative change in R2 = 2.8 %). After harmonization, patient data of all sites could be aggregated in a single analysis to infer the association between WMH volume and FA (R2 = 0.62), MD (R2 = 0.64), and PSMD (R2 = 0.60). CONCLUSIONS: We showed that RISH harmonization effectively removes acquisition-related differences in dMRI of elderly subjects while preserving sensitivity to SVD-related effects. This study provides proof of concept for future multicentre SVD studies with pooled datasets.

Effect of Fixed-Density Thresholding on Structural Brain Networks: A Demonstration in Cerebral Small Vessel Disease.

A popular solution to control for edge density variability in structural brain network analysis is to threshold the networks to a fixed density across all subjects. However, it remains unclear how this type of thresholding affects the basic network architecture in terms of edge weights, hub location, and hub connectivity and, especially, how it affects the sensitivity to detect disease-related abnormalities. We investigated these two questions in a cohort of patients with cerebral small vessel disease and age-matched controls. Brain networks were reconstructed from diffusion magnetic resonance imaging data using deterministic fiber tractography. Networks were thresholded to a fixed density by removing edges with the lowest number of streamlines. We compared edge length (mm), fractional anisotropy (FA), proportion of hub connections, and hub location between the unthresholded and the thresholded networks of each subject. Moreover, we compared weighted graph measures of global and local connectivity obtained from the (un)thresholded networks between patients and controls. We performed these analyses over a range of densities (2-20%). Results indicate that fixed-density thresholding disproportionally removes edges composed of long streamlines, but is independent of FA. The edges removed were not preferentially connected to hub or nonhub nodes. Over half of the original hubs were reproducible when networks were thresholded to a density ≥10%. Furthermore, the between-group differences in graph measures observed in the unthresholded network remained present after thresholding, irrespective of the chosen density. We therefore conclude that moderate fixed-density thresholds can successfully be applied to control for the effects of density in structural brain network analysis.