Automatic Denoising of Functional MRI Data: Combining Independent Component Analysis and Hierarchical Fusion of Classifiers

By Gholamreza Salimi Khorshidi, Gwenaëlle Douaud, Christian F. Beckmann, Matthew F. Glasser, Ludovica Griffanti, Stephen M. Smith. 2014

1,824 citations recorded by the source corpus.

Many sources of fluctuation contribute to the fMRI signal, and this makes identifying the effects that are truly related to the underlying neuronal activity difficult. Independent component analysis (ICA) - one of the most widely used techniques for the exploratory analysis of fMRI data - has shown to be a powerful technique in identifying various sources of neuronally-related and artefactual fluctuation in fMRI data (both with the application of external stimuli and with the subject “at rest”). ICA decomposes fMRI data into patterns of activity (a set of spatial maps and their corresponding time series) that are statistically independent and add linearly to explain voxel-wise time series. Given the set of ICA components, if the components representing “signal” (brain activity) can be distinguished form the “noise” components (effects of motion, non-neuronal physiology, scanner artefacts and other nuisance sources), the latter can then be removed from the data, providing an effective cleanup of structured noise. Manual classification of components is labour intensive and requires expertise; hence, a fully automatic noise detection algorithm that can reliably detect various types of noise sources (in both task and resting fMRI) is desirable. In this paper, we introduce FIX (“FMRIB’s ICA-based X-noiseifier”), which provides an automatic solution for denoising fMRI data via accurate classification of ICA components. For each ICA component FIX generates a large number of distinct spatial and temporal features, each describing a different aspect of the data (e.g., what proportion of temporal fluctuations are at high frequencies). The set of features is then fed into a multi-level classifier (built around several different Classifiers). Once trained through the hand-classification of a sufficient number of training datasets, the classifier can then automatically classify new datasets. The noise components can then be subtracted from (or regressed out of) the original data, to provide automated cleanup. On conventional resting-state fMRI (rfMRI) single-run datasets, FIX achieved about 95% overall accuracy. On high-quality rfMRI data from the Human Connectome Project, FIX achieves over 99% classification accuracy, and as a result is being used in the default rfMRI processing pipeline for generating HCP connectomes. FIX is publicly available as a plugin for FSL.

Subfields: Cognitive Neuroscience, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Signal Processing

Citations in the corpus

An overview of deep learning in medical imaging focusing on MRI (2018, methodology): “For example Bayesian Markov random field models [173], rough set theory [174], higherorder singular value decomposition [175], wavelets [176], independent component analysis [177], or higher order PDEs [178].”
ICA-based artifact removal diminishes scan site differences in multi-center resting-state fMRI (2015, methodology): “Previous studies using FIX have considered the theoretical and practical use of spatial ICA, classifier training and noise detection (Salimi-Khorshidi et al., 2014) and denoising (Griffanti et al., 2014).”
Data Quality Influences Observed Links Between Functional Connectivity and Behavior (2017, methodology): “FIX (FMRIB’s ICA-based X-noiseifier) is an automated IC-classifier designed to remove artifactual components from FC data (Salimi-Khorshidi et al. 2014).”
Hand classification of fMRI ICA noise components (2017, positive): “However, the final classification is usually reached through a consensus of multiple experts, especially regarding unknown components (when the raters do not or cannot agree, the component is labelled as unknown) (Griffanti et al., 2014; Rummel et al., 2013; Salimi-Khorshidi et al., 2014; Storti et ”
Image processing and Quality Control for the first 10,000 brain imaging datasets from UK Biobank (2017, methodology): “Probabilistic-tractography-based analysis In addition to the TBSS analyses, the preprocessed dMRI data is also fed into a tractography-based analysis.”
Mapping the human brain’s cortical-subcortical functional network organization (2018, positive): “Additionally , several of the regions included in this network were thoroughly investigated by Glasser et al. (201 6), establishing these regions as distinct functional entities with clear language functionali ty. Notably, the Power et al. (2011) partition (updated and visualized more fully by (Cole”
ICA-AROMA: A robust ICA-based strategy for removing motion artifacts from fMRI data (2015, positive): “In summary, we showed that spike regression, scrubbing, ICA-FIX and ICA-AROMA minimized the impact of head motion on functional connectivity estimates.”
Evaluation of ICA-AROMA and alternative strategies for motion artifact removal in resting state fMRI (2015, methodology): “A recent example is ‘ FMRIB's ICA-based X-noisei ﬁ er ’ (ICA-FIX) that implements noise component classi ﬁ cation using an extensive set of features and a multi-level classi-ﬁ er(Salimi-Khorshidietal.,2014).”
Global Signal Regression Strengthens Association between Resting-State Functional Connectivity and Behavior (2019, methodology): “In the case of the HCP dataset, it has previously been denoised using ICA-FIX (Salimi-Khorshidi et al., 2014; Griffanti et al., 2014).”
Spatial Topography of Individual-Specific Cortical Networks Predicts Human Cognition, Personality and Emotion (2017, methodology): “In the case of the HCP data, we utilized the MSMAll ICA-FIX data on fs_LR32K surface space (HCP S900 manual; Van Essen et al. 2012a; Van Essen et al. 2012b; Glasser et al. 2013; Smith et al. 2013; Griffanti et al., 2014; Salimi-Khorshidi et al. 2014) with additional nuisance regression, censoring (B”
Building a Science of Individual Differences from fMRI. (2016, methodology): “Automated algorithms, like ICA-FIX [108], rely on prior manual classification of components and are therefore not immune to this criticism.”
Mapping neurotransmitter systems to the structural and functional organization of the human neocortex (2021, methodology): “…include co-registration of the corrected images to the T1w structural MR images, brain extraction, normalization of whole brain intensity, high-pass filtering (>2,000s full width at half maximum (FWHM); to correct for scanner drifts) and removing additional noise using the ICA-FIX process 106,114 .”