The impact of diverse preprocessing pipelines on brain functional connectivity

Abstract

Abstract-Brain functional connectivity measured by functional magnetic resonance imaging was shown to be influenced by preprocessing procedures. We aim to describe this influence separately for different preprocessing factors and in 20 different most used preprocessing pipelines. We evaluate the effects of slice-timing correction and physiological noise filtering by RETROICOR, diverse levels of motion correction, and white matter, cerebrospinal fluid, and global signal filtering. With usage of three datasets, we show the impact on global metrics of restingstate functional brain networks and their reliability. We show negative effect of RETROICOR on reliability of metrics and disrupting effect of global signal regression on network topology. We do not support the use of slice-timing correction because it does not significantly influence any of the measured features. We also show that the selected types of preprocessing may affect averaged node strength, normalized clustering coefficient, normalized characteristic path length and modularity.

Neuropsychiatric disease classification using functional connectomics - results of the connectomics in neuroimaging transfer learning challenge

“…The effect of rsfMRI processing decisions on analysis is a large and important issue to address that is beyond the scope of our Challenge, but has been investigated by others (see e.g. Bowring et al (2019); Výtvarová et al (2017); Vergara et al (2017)). We note that a subset of the rsfMRI data has been released through ABIDE (Di Martino et al (2013)) and ADHD-200 9 .…”

Spatial-temporal-spectral EEG patterns of BOLD functional network connectivity dynamics

“…The dataset also contained simultaneously recorded ECG and respiration, which allowed us to pre-process fMRI data by RETROICOR. And finally, we tried many pre-processing strategies [79] to test their effect on functional connectivity estimation. For this study, we were able to select the best pipeline with the highest signal-noise separation (SNS) parameter [80].…”

Denoising Approach Affects Diagnostic Differences in Brain Connectivity across the Alzheimer’s Disease Continuum

“…Much work documenting the influence of denoising strategy on network topology has been conducted in young healthy adults (Aurich et al, 2015; Liang et al, 2012; Ciric et al, 2017; Vỳtvarová et al, 2017; Yan et al, 2013). A strength of the current investigation was that analyses were conducted in older adults at different stages of the AD continuum.…”

“…Here, we focus on the potential impact of preprocessing decisions, specifically denoising strategy, on AD diagnostic differences in graph metrics. An emerging literature provides evidence that the steps included and order of preprocessing pipelines affects the reliability of functional connectivity estimations and ultimately graph metrics (Alakörkkö, Saarimäki, Glerean, Saramäki, & Korhonen, 2017; Aurich, Alves Filho, Marques da Silva, & Franco, 2015; Borchardt et al, 2016; Liang et al, 2012; Ran et al, 2020; Vỳtvarová et al, 2017; Yan, Craddock, He, & Milham, 2013). For instance, Liang and colleagues (2012) found that preprocessing pipelines excluding global signal regression and including a wide bandpass frequency range (0.02-0.07) produced the most reliable topological estimates.…”