A method to mitigate spatio-temporally varying task-correlated motion artifacts from overt-speech fMRI paradigms in aphasia

Global Impact of Aging on the Hemodynamic Response Function in the Gray Matter of Human Cerebral Cortex

In functional magnetic resonance imaging, the hemodynamic response function (HRF) is a stereotypical response to local changes in cerebral hemodynamics and oxygen metabolism due to briefly (< 4 s) evoked neural activity. Accordingly, the HRF is often used as an impulse response with the assumption of linearity in data analysis. In cognitive aging studies, it has been very common to interpret differences in brain activation as age-related changes in neural activity. Contrary to this assumption, however, evidence has accrued that normal aging may also significantly affect the vasculature, thereby affecting cerebral hemodynamics and metabolism, confounding interpretation of fMRI cognitive aging studies. In this study, use was made of a multisensory task to evoke the HRF in ~87% of cerebral cortex in cognitively intact adults with ages ranging from 22 to 75 years. This widespread activation enabled us to investigate age trends in the spatial distributions of HRF characteristics within the majority of cortical gray matter, which we termed as global age trends. The task evoked both positive and negative HRFs, which were characterized using model-free parameters in native-space coordinates. We found significant global age trends in the distributions of HRF parameters in terms of both amplitudes (e.g., peak amplitude and contrast-to-noise ratio) and temporal dynamics (e.g., full-width-at-half-maximum). Our findings offer insight into how age-dependent changes affect neurovascular coupling and show promise for use of HRF parameters as non-invasive indicators for age-related pathology.

A Comparison of Denoising Approaches for Spoken Word Production Related Artefacts in Continuous Multiband fMRI Data

It is well-established from fMRI experiments employing gradient echo echo-planar imaging (EPI) sequences that overt speech production introduces signal artefacts compromising accurate detection of task-related responses. Both design and post-processing (denoising) techniques have been proposed and implemented over the years to mitigate the various noise sources. Recently, fMRI studies of speech production have begun to adopt multiband EPI sequences that offer better signal-to-noise ratio (SNR) and temporal resolution allowing adequate sampling of physiological noise sources (e.g., respiration, cardiovascular effects) and reduced scanner acoustic noise. However, these new sequences may also introduce additional noise sources. In this study, we demonstrate the impact of applying several noise-estimation and removal approaches to continuous multiband fMRI data acquired during a naming-to-definition task, including rigid body motion regression and outlier censoring, principal component analysis for removal of cerebrospinal fluid (CSF)/edge-related noise components, and global fMRI signal regression (using two different approaches) compared to a baseline of realignment and unwarping alone. Our results show the strongest and most spatially extensive sources of physiological noise are the global signal fluctuations arising from respiration and muscle action and CSF/edge-related noise components, with residual rigid body motion contributing relatively little variance. Interestingly, denoising approaches tended to reduce and enhance task-related BOLD signal increases and decreases, respectively. Global signal regression using a voxel-wise linear model of the global signal estimated from unmasked data resulted in dramatic improvements in temporal SNR. Overall, these findings show the benefits of combining continuous multiband EPI sequences and denoising approaches to investigate the neurobiology of speech production.

Bayesian longitudinal tensor response regression for modeling neuroplasticity

A major interest in longitudinal neuroimaging studies involves investigating voxel-level neuroplasticity due to treatment and other factors across visits. However, traditional voxel-wise methods are beset with several pitfalls, which can compromise the accuracy of these approaches. We propose a novel Bayesian tensor response regression approach for longitudinal imaging data, which pools information across spatially distributed voxels to infer significant changes while adjusting for covariates. The proposed method, which is implemented using Markov chain Monte Carlo (MCMC) sampling, utilizes low-rank decomposition to reduce dimensionality and preserve spatial configurations of voxels when estimating coefficients. It also enables feature selection via joint credible regions which respect the shape of the posterior distributions for more accurate inference. In addition to group level inferences, the method is able to infer individual-level neuroplasticity, allowing for examination of personalized disease or recovery trajectories. The advantages of the proposed approach in terms of prediction and feature selection over voxel-wise regression are highlighted via extensive simulation studies. Subsequently, we apply the approach to a longitudinal Aphasia dataset consisting of task functional MRI images from a group of subjects who were administered either a control intervention or intention treatment at baseline and were followed up over subsequent visits. Our analysis revealed that while the control therapy showed long-term increases in brain activity, the intention treatment produced predominantly short-term changes, both of which were concentrated in distinct localized regions. In contrast, the voxel-wise regression failed to detect any significant neuroplasticity after multiplicity adjustments, which is biologically implausible and implies lack of power.

Methodologies for task-fMRI based prognostic biomarkers in response to aphasia treatment

With the diversity in aphasia coupled with diminished gains at the chronic phase, it is imperative to deliver effective rehabilitation plans. Treatment outcomes have therefore been predicted using lesion-to-symptom mapping, but this method lacks holistic functional information about the language-network. This study, therefore, aims to develop whole-brain task-fMRI multivariate analysis to neurobiologically inspect lesion impacts on the language-network and predict behavioral outcomes in persons with aphasia (PWA) undergoing language therapy. In 14 chronic PWA, semantic fluency task-fMRI and behavioral measures were collected to develop prediction methodologies for post-treatment outcomes. Then, a recently developed imaging-based multivariate method to predict behavior (i.e., LESYMAP) was optimized to intake whole-brain task-fMRI data, and systematically tested for reliability with mass univariate methods. We also accounted for lesion size in both methods. Results showed that both mass univariate and multivariate methods identified unique biomarkers for semantic fluency improvements from baseline to 2-weeks post-treatment. Additionally, both methods demonstrated reliable spatial overlap in task-specific areas including the right middle frontal gyrus when identifying biomarkers of language discourse. Thus whole-brain task-fMRI multivariate analysis has the potential to identify functionally meaningful prognostic biomarkers even for relatively small sample sizes. In sum, our task-fMRI based multivariate approach holistically estimates post-treatment response for both word and sentence production and may serve as a complementary tool to mass univariate analysis in developing brain-behavior relationships for improved personalization of aphasia rehabilitation regimens.

Not All Lesioned Tissue Is Equal: Identifying Pericavitational Areas in Chronic Stroke With Tissue Integrity Gradation via T2w T1w Ratio

Stroke-related tissue damage within lesioned brain areas is topologically non-uniform and has underlying tissue composition changes that may have important implications for rehabilitation. However, we know of no uniformly accepted, objective non-invasive methodology to identify pericavitational areas within the chronic stroke lesion. To fill this gap, we propose a novel magnetic resonance imaging (MRI) methodology to objectively quantify the lesion core and surrounding pericavitational perimeter, which we call tissue integrity gradation via T2w T1w ratio (TIGR). TIGR uses standard T1-weighted (T1w) and T2-weighted (T2w) anatomical images routinely collected in the clinical setting. TIGR maps are analyzed with relation to subject-specific gray matter and cerebrospinal fluid thresholds and binned to create a false colormap of tissue damage within the stroke lesion, and these are further categorized into low-, medium-, and high-damage areas. We validate TIGR by showing that the cerebral blood flow within the lesion reduces with greater tissue damage (p = 0.005). We further show that a significant task activity can be detected in pericavitational areas and that medium-damage areas contain a significantly lower magnitude of hemodynamic response function than the adjacent damaged areas (p < 0.0001). We also demonstrate the feasibility of using TIGR maps to extract multivariate brain-behavior relationships (p < 0.05) and show general agreement in location compared to binary lesion, T1w-only, and T2w-only maps but that the extent of brain behavior maps may depend on signal sensitivity as denoted by the sparseness coefficient (p < 0.0001). Finally, we show the feasibility of quantifying TIGR in early and late subacute stroke phases, where higher-damage areas were smaller in size (p = 0.002) and that lesioned voxels transition from lower to higher damage with increasing time post-stroke (p = 0.004). We conclude that TIGR is able to (1) identify tissue damage gradient within the stroke lesion across different post-stroke timepoints and (2) more objectively delineate lesion core from pericavitational areas wherein such areas demonstrate reasonable and expected physiological and functional impairments. Importantly, because T1w and T2w scans are routinely collected in the clinic, TIGR maps can be readily incorporated in clinical settings without additional imaging costs or patient burden to facilitate decision processes related to rehabilitation planning.