Comparison of multi-subject ICA methods for analysis of fMRI data

By Erik B. Erhardt, Srinivas Rachakonda, Edward J. Bedrick, Elena A. Allen, T. Adalı, Vince D. Calhoun. Human Brain Mapping, 2010

718 citations recorded by the source corpus.

Spatial independent component analysis (ICA) applied to functional magnetic resonance imaging (fMRI) data identifies functionally connected networks by estimating spatially independent patterns from their linearly mixed fMRI signals. Several multi‐subject ICA approaches estimating subject‐specific time courses (TCs) and spatial maps (SMs) have been developed, however, there has not yet been a full comparison of the implications of their use. Here, we provide extensive comparisons of four multi‐subject ICA approaches in combination with data reduction methods for simulated and fMRI task data. For multi‐subject ICA, the data first undergo reduction at the subject and group levels using principal component analysis (PCA). Comparisons of subject‐specific, spatial concatenation, and group data mean subject‐level reduction strategies using PCA and probabilistic PCA (PPCA) show that computationally intensive PPCA is equivalent to PCA, and that subject‐specific and group data mean subject‐level PCA are preferred because of well‐estimated TCs and SMs. Second, aggregate independent components are estimated using either noise‐free ICA or probabilistic ICA (PICA). Third, subject‐specific SMs and TCs are estimated using back‐reconstruction. We compare several direct group ICA (GICA) back‐reconstruction approaches (GICA1‐GICA3) and an indirect back‐reconstruction approach, spatio‐temporal regression (STR, or dual regression). Results show the earlier group ICA (GICA1) approximates STR, however STR has contradictory assumptions and may show mixed‐component artifacts in estimated SMs. Our evidence‐based recommendation is to use GICA3, introduced here, with subject‐specific PCA and noise‐free ICA, providing the most robust and accurate estimated SMs and TCs in addition to offering an intuitive interpretation. Hum Brain Mapp, 2011. © 2010 Wiley Periodicals, Inc.

Subfields: Cognitive Neuroscience, Signal Processing

View Paper: https://www.semanticscholar.org/paper/a9054e09b981bccbe4f58366abe9606b6f33f26e

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

Impact of In-Scanner Head Motion on Multiple Measures of Functional Connectivity: Relevance for Studies of Neurodevelopment in Youth (2012, methodology): “…the effects of motion on dualregression ICA, it should be noted that other methods exist for estimating component strength in individual subjects (Erhardt et al., 2010); future work is necessary to evaluate the relative susceptibility of different methods to the confounding effects of motion.”
Dynamic functional connectivity analysis reveals transient states of dysconnectivity in schizophrenia (2014, methodology): “Resting state fMRI scans were acquired using a standard gradient-echo echo planar imaging paradigm: FOV of 220 × 220 mm (64 × 64 matrix), TR= 2 s, TE= 30 ms, FA= 770, 162 volumes, 32 sequential ascending axial slices of 4 mm thickness and 1 mm skip.”
Capturing inter-subject variability with group independent component analysis of fMRI data: A simulation study (2012, methodology): “We retain T1=T−1 principal components (PCs) since a relatively large number of subject-specific PCs has been shown to improve reconstruction of subject components (Erhardt et al., 2011b).”
Tracking whole-brain connectivity dynamics in the resting state. (2014, positive): “Subject-specific SMs (Si) and time courses (TCs, Ri) were estimated using the GICA1 backreconstruction method based on PCA compression and projection (Calhoun et al. 2001; Erhardt et al. 2011), which is equivalent to least-squares based dual regression (Filippini et al.”
Using Dual Regression to Investigate Network Shape and Amplitude in Functional Connectivity Analyses (2017, background): “As such, while back-projection approaches may be tuneable to give good results in empirical studies using speciﬁc conditions in simulated data (Erhardt et al., 2011; Allen et al., 2012), we advocate the dual regression approach as being theoretically well-principled and practically “safer.””
Multi-subject Independent Component Analysis of fMRI: A Decade of Intrinsic Networks, Default Mode, and Neurodiagnostic Discovery (2012, neutral): “These various approaches yield largely similar results, though memory requirements vary considerably [44].”
A Review of Feature Reduction Techniques in Neuroimaging (2013, background): “The group-level ICA back-reconstruction step is explored elsewhere (Erhardt et al. 2011) .”
CLASSIFICATION OF SCHIZOPHRENIA AND BIPOLAR PATIENTS USING STATIC AND DYNAMIC RESTING-STATE FMRI BRAIN CONNECTIVITY (2016, methodology): “Subject-specific spatial maps (SMs) and time-courses (TCs) were estimated using the GICA1 back-reconstruction method based on PCA compression and projection (Calhoun et al., 2001b; Erhardt et al., 2011).”
Group information guided ICA for fMRI data analysis (2013, neutral): “14 (continued). than GICA1, consistent with the finding reported in Erhardt et al. (2011).”
Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion (2011, neutral): “ICA decomposition or regression of motion estimates, to remove spurious motion-related signal from the data (Beckmann and Smith, 2005; Erhardt et al., 2010; Fox et al., 2009; Robinson et al., 2009a; Weissenbacher et al., 2009).”
A Baseline for the Multivariate Comparison of Resting-State Networks (2011, methodology): “, 2009), evidence suggests that noise-free PCA with noise-free ICA using GICA3 provides more robust results with a more intuitive and natural interpretation (Erhardt et al., 2010).”
Functional network organization of the human brain (2011, result): “Integrating the Present Findings with Other Approaches to Whole-Brain rs-fcMRI Analysis Recent investigations into the structure of functional brain organization using a variety of methods (Erhardt et al., 2010; Yeo et al., 2011) have found some similar (but not identical) sets of resting state netw”