New analysis method for functional brain imaging: White noise removed T2* variation mapping using multi-echo EPI

BACKGROUND

Generally, the analysis of functional magnetic resonance imaging (fMRI) using echo-planar imaging (EPI) data is based on independent component analysis (ICA) and the general linear model (GLM). The application of these two approaches is highly independent, like GLM is for task-related activation mapping, and ICA is for resting-state imaging. Herein, we propose white noise-removed T₂*-variation mapping as a new analysis method for fMRI that integrates the two conventional mapping approaches.

NEW METHOD

We derived the standard deviation to the mean-square ratio of the true T₂* signal from the multi-echo EPI (ME-EPI) dataset. For the true T₂*-variation-based value, we removed the S₀ (initial signal intensity) and white noise component from the variation in the EPI signal using signal-coherence analysis of the echo time (TE) dataset and slope analysis of the TE-variated coefficient of variation of the ME-EPI dataset.

RESULTS

The activation mapping for a visual task and resting-state imaging by the proposed method showed the reliable activation map in the visual cortex area and area for the typical default mode network, with white noise and the S₀ component removed.

COMPARISON WITH EXISTING METHODS

Conventional analyses for fMRI cannot be applied to both activation mapping and resting-state imaging, with white noise removed, while the proposed method can be applied.

CONCLUSIONS

We demonstrated white noise-removed true T₂*-variation-based mapping as a new functional brain analysis approach. We expect the method allows studying in which that the association between task timing and brain activity is somewhat uncertain, such as studies of emotion and awareness.

Improving effect size estimation and statistical power with multi-echo fMRI and its impact on understanding the neural systems supporting mentalizing

Functional magnetic resonance imaging (fMRI) research is routinely criticized for being statistically underpowered due to characteristically small sample sizes and much larger sample sizes are being increasingly recommended. Additionally, various sources of artifact inherent in fMRI data can have detrimental impact on effect size estimates and statistical power. Here we show how specific removal of non-BOLD artifacts can improve effect size estimation and statistical power in task-fMRI contexts, with particular application to the social-cognitive domain of mentalizing/theory of mind. Non-BOLD variability identification and removal is achieved in a biophysical and statistically principled manner by combining multi-echo fMRI acquisition and independent components analysis (ME-ICA). Without smoothing, group-level effect size estimates on two different mentalizing tasks were enhanced by ME-ICA at a median rate of 24% in regions canonically associated with mentalizing, while much more substantial boosts (40-149%) were observed in non-canonical cerebellar areas. Effect size boosting occurs via reduction of non-BOLD noise at the subject-level and consequent reductions in between-subject variance at the group-level. Smoothing can attenuate ME-ICA-related effect size improvements in certain circumstances. Power simulations demonstrate that ME-ICA-related effect size enhancements enable much higher-powered studies at traditional sample sizes. Cerebellar effects observed after applying ME-ICA may be unobservable with conventional imaging at traditional sample sizes. Thus, ME-ICA allows for principled design-agnostic non-BOLD artifact removal that can substantially improve effect size estimates and statistical power in task-fMRI contexts. ME-ICA could mitigate some issues regarding statistical power in fMRI studies and enable novel discovery of aspects of brain organization that are currently under-appreciated and not well understood.

The quest for the best: The impact of different EPI sequences on the sensitivity of random effect fMRI group analyses

We compared the sensitivity of standard single-shot 2D echo planar imaging (EPI) to three advanced EPI sequences, i.e., 2D multi-echo EPI, 3D high resolution EPI and 3D dual-echo fast EPI in fixed effect and random effects group level fMRI analyses at 3 T. The study focused on how well the variance reduction in fixed effect analyses achieved by advanced EPI sequences translates into increased sensitivity in the random effects group level analysis. The sensitivity was estimated in a functional MRI experiment of an emotional learning and a reward based learning tasks in a group of 24 volunteers. Each experiment was acquired with the four different sequences. The task-related response amplitude, contrast level and respective t-value were proxies for the functional sensitivity across the brain. All three advanced EPI methods increased the sensitivity in the fixed effects analyses, but standard single-shot 2D EPI provided a comparable performance in random effects group analysis when whole brain coverage and moderate resolution are required. In this experiment inter-subject variability determined the sensitivity of the random effects analysis for most brain regions, making the impact of EPI pulse sequence improvements less relevant or even negligible for random effects analyses. An exception concerns the optimization of EPI reducing susceptibility-related signal loss that translates into an enhanced sensitivity e.g. in the orbitofrontal cortex for multi-echo EPI. Thus, future optimization strategies may best aim at reducing inter-subject variability for higher sensitivity in standard fMRI group studies at moderate spatial resolution.

•

We compared the sensitivity of standard and three advanced EPI sequences in group level fMRI analysis.

•

Advanced EPI methods increased the sensitivity in the fixed effect group analyses.

•

All EPI methods provided similar sensitivity in random effect group analysis.

•

Inter-subject variance determined the sensitivity of the random effect analysis.

•

Future sequence optimization best aim at reducing inter-subject rather then intra-subject variance.

BOLD sensitivity and SNR characteristics of parallel imaging-accelerated single-shot multi-echo EPI for fMRI

Echo-planar imaging (EPI) is a standard procedure in functional magnetic resonance imaging (fMRI) for measuring changes in the blood oxygen level-dependent (BOLD) signal associated with neuronal activity. The images obtained from fMRI with EPI, however, exhibit signal dropouts and geometric distortions. Parallel imaging (PI), due to its short readout, accelerates image acquisition and might reduce dephasing in phase-encoding direction. The concomitant loss of signal-to-noise ratio (SNR) might be compensated through single-shot multi-echo EPI (mEPI). We systematically compared the temporal SNR and BOLD sensitivity of single echoes (TE=15, 45, and 75ms) and contrast-optimized mEPI with and without PI and mEPI-based denoising. Audio-visual stimulation under natural viewing conditions activated distributed neural networks. Heterogeneous SNR, noise gain, and sensitivity maps emerged. In single echoes, SNR and BOLD sensitivity followed the predicted dependency on echo time (TE) and were reduced under PI. However, the combination of echoes with mEPI recovered the quality parameters and increased BOLD signal changes at circumscribed fronto-polar and deep brain structures. We suggest applying PI only in combination with mEPI to reduce imaging artifacts and conserve BOLD sensitivity.