Robust data driven model order estimation for independent component analysis of FMRI data with low contrast to noise.

Waqas Majeed,Malcolm J Avison

doi:10.1371/journal.pone.0094943

Abstract

Independent component analysis (ICA) has been successfully utilized for analysis of functional MRI (fMRI) data for task related as well as resting state studies. Although it holds the promise of becoming an unbiased data-driven analysis technique, a few choices have to be made prior to performing ICA, selection of a method for determining the number of independent components (nIC) being one of them. Choice of nIC has been shown to influence the ICA maps, and various approaches (mostly relying on information theoretic criteria) have been proposed and implemented in commonly used ICA analysis packages, such as MELODIC and GIFT. However, there has been no consensus on the optimal method for nIC selection, and many studies utilize arbitrarily chosen values for nIC. Accurate and reliable determination of true nIC is especially important in the setting where the signals of interest contribute only a small fraction of the total variance, i.e. very low contrast-to-noise ratio (CNR), and/or very focal response. In this study, we evaluate the performance of different model order selection criteria and demonstrate that the model order selected based upon bootstrap stability of principal components yields more reliable and accurate estimates of model order. We then demonstrate the utility of this fully data-driven approach to detect weak and focal stimulus-driven responses in real data. Finally, we compare the performance of different multi-run ICA approaches using pseudo-real data.

Highlights

The inherent complexity of functional neuroimaging data has led to an increased interest in analysis techniques capable of revealing the intrinsic architecture of the data
We evaluated the performance of different datadriven number of independent components (nIC) estimation approaches for simulated data, demonstrating that the model order estimated using bootstrap stability analysis (BSA) of principal components [13] provided a better estimate of the true number of components over a wide range of contrast-to-noise ratio (CNR) and preprocessing conditions
The temporal profiles associated with the stimulus-related independent component obtained using ICA on averaged runs (ICAavg) showed greater correlation with the stimulus time-course, compared with that corresponding to ICA with 2-step PCA reduction (ICAcat) (0.6160.07 vs 0.5160.17). These findings suggest that greater functional contrast and spatial/temporal accuracy can be achieved when ICAavg is used, where spatial/temporal accuracy is estimated by similarity of the maps/ time-courses with those associated with hypothesis-driven analysis

Summary

Introduction

The inherent complexity of functional neuroimaging data has led to an increased interest in analysis techniques capable of revealing the intrinsic architecture of the data. The ability to partition the spatiotemporal variation according to the underlying sources would be the ideal way of isolating the neural activity (or any other components of interest) and developing an in-depth understanding of the functional architecture. An implicit assumption of ICA is that the number of observations is equal to the number of underlying sources (or alternatively, the true number of sources is known, or can be estimated). This is an important assumption that strongly influences the ICA result, and will be discussed later in the text

Methods

Results

Conclusion