Overcoming the effects of false positives and threshold bias in graph theoretical analyses of neuroimaging data

Abstract

Graph theory (GT) is a powerful framework for quantifying topological features of neuroimaging-derived functional and structural networks. However, false positive (FP) connections arise frequently and influence the inferred topology of networks. Thresholding is often used to overcome this problem, but an appropriate threshold often relies on a priori assumptions, which will alter inferred network topologies.Four common network metrics (global efficiency, mean clustering coefficient, mean betweenness and smallworldness) were tested using a model tractography dataset. It was found that all four network metrics were significantly affected even by just one FP. Results also show that thresholding effectively dampens the impact of FPs, but at the expense of adding significant bias to network metrics.In a larger number (n = 248) of tractography datasets, statistics were computed across random group permutations for a range of thresholds, revealing that statistics for network metrics varied significantly more than for non-network metrics (i.e., number of streamlines and number of edges). Varying degrees of network atrophy were introduced artificially to half the datasets, to test sensitivity to genuine group differences. For some network metrics, this atrophy was detected as significant (p < 0.05, determined using permutation testing) only across a limited range of thresholds.We propose a multi-threshold permutation correction (MTPC) method, based on the cluster-enhanced permutation correction approach, to identify sustained significant effects across clusters of thresholds. This approach minimises requirements to determine a single threshold a priori. We demonstrate improved sensitivity of MTPC-corrected metrics to genuine group effects compared to an existing approach and demonstrate the use of MTPC on a previously published network analysis of tractography data derived from a clinical population.In conclusion, we show that there are large biases and instability induced by thresholding, making statistical comparisons of network metrics difficult. However, by testing for effects across multiple thresholds using MTPC, true group differences can be robustly identified.

Enhanced estimations of post‐stroke aphasia severity using stacked multimodal predictions

“…In this study we used a novel approach to combine multithreshold graph theory measures. The identification of appropriate thresholds when constructing graph theory measures is an ongoing challenge that has led some authors to propose the use of multithreshold graph measures [Drakesmith et al, ]. Our approach, which relies on multithreshold prediction stacking, may offer an alternative solution to the search for threshold‐free graph theory analyses.…”

Graph theory analysis of resting‐state functional magnetic resonance imaging in essential tremor

“…The thresholding method has proved capable of reducing the false positives resulting of the comparison of low correlated nodes between groups (Drakesmith et al, 2015). The scrubbing regressors, provided by CompCor, Artifact Removal Tool, and the motion parameters estimated for each participant, were included as covariates.…”

“…An option is to apply a correction over the graph prior to network analysis, thresholding the connectivity matrixes. The thresholding method has proved capable of reducing the false positives resulting of the comparison of low correlated nodes between groups (Drakesmith et al, 2015). For this reason, the correlation previously obtained for all participants were binarized, thresholding them over increasing correlation values in steps of 0.1, starting from 0.1.…”

Stability and sensitivity of structural connectomes: effect of thresholding and filtering and demonstration in neurodegeneration

“…However currently there is no agreement in the literature regarding the best practice for threshold implementation (Qi et al, 2015). An absolute threshold defines a value, below which connections are removed (Figure 1, top row) (Daianu et al, 2015;Drakesmith et al, 2015;Li et al, 2016). A relative threshold retains a defined proportion of the strongest connections in the network (Figure 1, bottom row) (Mueller et al, 2015;Yao et al, 2010).…”