Template based rotation: a method for functional connectivity analysis with a priori templates

Comparing data-driven physiological denoising approaches for resting-state fMRI: implications for the study of aging

Introduction

Physiological nuisance contributions by cardiac and respiratory signals have a significant impact on resting-state fMRI data quality. As these physiological signals are often not recorded, data-driven denoising methods are commonly used to estimate and remove physiological noise from fMRI data. To investigate the efficacy of these denoising methods, one of the first steps is to accurately capture the cardiac and respiratory signals, which requires acquiring fMRI data with high temporal resolution.

Methods

In this study, we used such high-temporal resolution fMRI data to evaluate the effectiveness of several data-driven denoising methods, including global-signal regression (GSR), white matter and cerebrospinal fluid regression (WM-CSF), anatomical (aCompCor) and temporal CompCor (tCompCor), ICA-AROMA. Our analysis focused on the consequence of changes in low-frequency, cardiac and respiratory signal power, as well as age-related differences in terms of functional connectivity (fcMRI).

Results

Our results confirm that the ICA-AROMA and GSR removed the most physiological noise but also more low-frequency signals. These methods are also associated with substantially lower age-related fcMRI differences. On the other hand, aCompCor and tCompCor appear to be better at removing high-frequency physiological signals but not low-frequency signal power. These methods are also associated with relatively higher age-related fcMRI differences, whether driven by neuronal signal or residual artifact. These results were reproduced in data downsampled to represent conventional fMRI sampling frequency. Lastly, methods differ in performance depending on the age group.

Discussion

While this study cautions direct comparisons of fcMRI results based on different denoising methods in the study of aging, it also enhances the understanding of different denoising methods in broader fcMRI applications.

Analysis of Resting-State Functional Magnetic Resonance Imaging in Alzheimer's Disease

Alzheimer's disease (AD) has been characterized by widespread network disconnection among brain regions, widely overlapping with the hallmarks of the disease. Functional connectivity has been studied with an upward trend in the last two decades, predominantly in AD among other neuropsychiatric disorders, and presents a potential biomarker with various features that might provide unique contributions to foster our understanding of neural mechanisms of AD. The resting-state functional MRI (rs-fMRI) is usually used to measure the blood-oxygen-level-dependent signals that reflect the brain's functional connectivity. Nevertheless, the rs-fMRI is still underutilized, which might be due to the fairly complex acquisition and analytic methodology. In this chapter, we presented the common methods that have been applied in rs-fMRI literature, focusing on the studies on individuals in the continuum of AD. The key methodological aspects will be addressed that comprise acquiring, processing, and interpreting rs-fMRI data. More, we discussed the current and potential implications of rs-fMRI in AD.

Physical activity is associated with increased resting-state functional connectivity in networks predictive of cognitive decline in clinically unimpaired older adults

Introduction

Physical activity (PA) promotes resilience with respect to cognitive decline, although the underlying mechanisms are not well understood. We examined the associations between objectively measured PA and resting-state functional connectivity magnetic resonance imaging (rs-fcMRI) across seven anatomically distributed neural networks.

Methods

rs-fcMRI, amyloid beta (Aβ) positron emission tomography (PET), PA (steps/day × 1 week), and longitudinal cognitive (Preclinical Alzheimer's Cognitive Composite) data from 167 cognitively unimpaired adults (ages 63 to 90) were used. We used linear and linear mixed-effects regression models to examine the associations between baseline PA and baseline network connectivity and between PA, network connectivity, and longitudinal cognitive performance.

Results

Higher PA was associated selectively with greater connectivity in three networks previously associated with cognitive decline (default, salience, left control). This association with network connectivity accounted for a modest portion of PA's effects on Aβ-related cognitive decline.

Discussion

Although other mechanisms are likely present, PA may promote resilience with respect to Aß-related cognitive decline, partly by increasing connectivity in a subset of cognitive networks.

Performance of Temporal and Spatial Independent Component Analysis in Identifying and Removing Low-Frequency Physiological and Motion Effects in Resting-State fMRI

Effective separation of signal from noise (including physiological processes and head motion) is one of the chief challenges for improving the sensitivity and specificity of resting-state fMRI (rs-fMRI) measurements and has a profound impact when these noise sources vary between populations. Independent component analysis (ICA) is an approach for addressing these challenges. Conventionally, due to the lower amount of temporal than spatial information in rs-fMRI data, spatial ICA (sICA) is the method of choice. However, with recent developments in accelerated fMRI acquisitions, the temporal information is becoming enriched to the point that the temporal ICA (tICA) has become more feasible. This is particularly relevant as physiological processes and motion exhibit very different spatial and temporal characteristics when it comes to rs-fMRI applications, leading us to conduct a comparison of the performance of sICA and tICA in addressing these types of noise. In this study, we embrace the novel practice of using theory (simulations) to guide our interpretation of empirical data. We find empirically that sICA can identify more noise-related signal components than tICA. However, on the merit of functional-connectivity results, we find that while sICA is more adept at reducing whole-brain motion effects, tICA performs better in dealing with physiological effects. These interpretations are corroborated by our simulation results. The overall message of this study is that if ICA denoising is to be used for rs-fMRI, there is merit in considering a hybrid approach in which physiological and motion-related noise are each corrected for using their respective best-suited ICA approach.

Networks of worry-towards a connectivity-based signature of late-life worry using higher criticism

Severe worry is a complex transdiagnostic phenotype independently associated with increased morbidity, including cognitive impairment and cardiovascular diseases. We investigated the neurobiological basis of worry in older adults by analyzing resting state fMRI using a large-scale network-based approach. We collected resting fMRI on 77 participants (>50 years old) with varying worry severity. We computed region-wise connectivity across the default mode network (DMN), anterior salience network, and left executive control network. All 22,366 correlations were regressed on worry severity and adjusted for age, sex, race, education, disease burden, depression, anxiety, rumination, and neuroticism. We employed higher criticism, a second-level method of significance testing for rare and weak features, to reveal the functional connectivity patterns associated with worry. The analysis suggests that worry has a complex, yet distinct signature associated with resting state functional connectivity. Intra-connectivities and inter-connectivities of the DMN comprise the dominant contribution. The anterior cingulate, temporal lobe, and thalamus are heavily represented with overwhelmingly negative association with worry. The prefrontal regions are also strongly represented with a mix of positive and negative associations with worry. Identifying the most salient connections may be useful for targeted interventions for reducing morbidity associated with severe worry in older adults.

Multiple bouts of high-intensity interval exercise reverse age-related functional connectivity disruptions without affecting motor learning in older adults

Exercise has emerged as an intervention that may mitigate age-related resting state functional connectivity and sensorimotor decline. Here, 42 healthy older adults rested or completed 3 sets of high-intensity interval exercise for a total of 23 min, then immediately practiced an implicit motor task with their non-dominant hand across five separate sessions. Participants completed resting state functional MRI before the first and after the fifth day of practice; they also returned 24-h and 35-days later to assess short- and long-term retention. Independent component analysis of resting state functional MRI revealed increased connectivity in the frontoparietal, the dorsal attentional, and cerebellar networks in the exercise group relative to the rest group. Seed-based analysis showed strengthened connectivity between the limbic system and right cerebellum, and between the right cerebellum and bilateral middle temporal gyri in the exercise group. There was no motor learning advantage for the exercise group. Our data suggest that exercise paired with an implicit motor learning task in older adults can augment resting state functional connectivity without enhancing behaviour beyond that stimulated by skilled motor practice.

Neural Correlates of Group Versus Individual Problem Solving Revealed by fMRI

Group problem solving is a prototypical complex collective intellectual activity. Psychological research provides compelling evidence that problem solving in groups is both qualitatively and quantitatively different from doing so alone. However, the question of whether individual and collective problem solving involve the same neural substrate has not yet been addressed, mainly due to methodological limitations. In the current study, functional magnetic resonance imaging was performed to compare brain activation when participants solved Raven-like matrix problems in a small group and individually. In the group condition, the participant in the scanner was able to discuss the problem with other team members using a special communication device. In the individual condition, the participant was required to think aloud while solving the problem in the silent presence of the other team members. Greater activation was found in several brain regions during group problem solving, including the medial prefrontal cortex; lateral parietal, cingulate, precuneus and retrosplenial cortices; frontal and temporal poles. These areas have been identified as potential components of the so-called "social brain" on the basis of research using offline judgments of material related to socializing. Therefore, this study demonstrated the actual involvement of these regions in real-time social interactions, such as group problem solving. However, further connectivity analysis revealed that the social brain components are co-activated, but do not increase their coupling during cooperation as would be suggested for a holistic network. We suggest that the social mode of the brain may be described instead as a re-configuration of connectivity between basic networks, and we found decreased connectivity between the language and salience networks in the group compared to the individual condition. A control experiment showed that the findings from the main experiment cannot be entirely accounted for by discourse comprehension. Thus, the study demonstrates affordances provided by the presented new technique for neuroimaging the "group mind," implementing the single-brain version of the second-person neuroscience approach.

Resting-state functional connectivity and amyloid burden influence longitudinal cortical thinning in the default mode network in preclinical Alzheimer's disease

Proteinopathies are key elements in the pathogenesis of age-related neurodegenerative diseases, particularly Alzheimer's disease (AD), with the nature and location of the proteinopathy characterizing much of the disease phenotype. Susceptibility of brain regions to pathology may partly be determined by intrinsic network structure and connectivity. It remains unknown, however, how these networks inform the disease cascade in the context of AD biomarkers, such as beta-amyloid (Aβ), in clinically-normal older adults.The default-mode network (DMN), a prominent intrinsic network, is heavily implicated in AD due to its spatial overlap with AD atrophy patterns and tau deposition. We investigated the influence of baseline Aβ positron emission tomography (PET) signal and intrinsic DMN connectivity on DMN-specific cortical thinning in 120 clinically-normal older adults from the Harvard Aging Brain Study (73 ± 6 years, 58% Female, CDR = 0). Participants underwent11C Pittsburgh Compound-B (PiB) PET, 18F flortaucipir (FTP) PET, and resting-state MRI scans at baselineand longitudinal MRI (3.6 ± 0.96 scans; 5.04 ± 0.8 years). Linear mixed models tested relationships between baseline PiB and DMN connectivity on cortical thinning in a composite of DMN regions. Lower DMN connectivity was associated with faster cortical thinning, but only in those with elevated baseline PiB-PET signal. This relationship was network specific, in that the frontoparietal control network did not account for the observed association. Additionally, the relationship was independent of inferior temporal lobe FTP-PET signal. Our findings provide evidence that compromised DMN connectivity, in the context of preclinical AD, foreshadows neurodegeneration in DMN regions.

Multiple markers contribute to risk of progression from normal to mild cognitive impairment

OBJECTIVE

To identify a parsimonious set of markers that optimally predicts subsequent clinical progression from normal to mild cognitive impairment (MCI).

METHODS

250 clinically normal adults (mean age = 73.6 years, SD = 6.0) from the Harvard Aging Brain Study were assessed at baseline on a wide set of markers, including magnetic resonance imaging markers of gray matter thickness and volume, white matter lesions, fractional anisotropy, resting state functional connectivity, positron emission tomography markers of glucose metabolism and β-amyloid (Aβ) burden, and a measure of vascular risk. Participants were also tested annually on a battery of clinical and cognitive tests (median follow-up = 5.0 years, SD = 1.66). We applied least absolute shrinkage and selection operator (LASSO) Cox models to determine the minimum set of non-redundant markers that predicts subsequent clinical progression from normal to MCI, adjusting for age, sex, and education.

RESULTS

23 participants (9.2%) progressed to MCI over the study period (mean years of follow-up to diagnosis = 3.96, SD = 1.89). Progression was predicted by several brain markers, including reduced entorhinal thickness (hazard ratio, HR = 1.73), greater Aβ burden (HR = 1.58), lower default network connectivity (HR = 1.42), and smaller hippocampal volume (HR = 1.30). When cognitive test scores were added to the model, the aforementioned neuroimaging markers remained significant and lower striatum volume as well as lower scores on baseline memory and processing speed tests additionally contributed to progression.

CONCLUSION

Among a large set of brain, vascular and cognitive markers, a subset of markers independently predicted progression from normal to MCI. These markers may enhance risk stratification by identifying clinically normal individuals who are most likely to develop clinical symptoms and would likely benefit most from therapeutic intervention.

Personalizing neuromodulation

In the era of "big data", we are gaining rich person-specific information about neuroanatomy, neural function, and cognitive functions. However, the optimal ways to create precise approaches to optimize individuals' mental functions in health and disease are unclear. Multimodal analysis and modeling approaches can guide neuromodulation by combining anatomical networks, functional signal analysis, and cognitive neuroscience paradigms in single subjects. Our progress could be improved by progressing from statistical fits to mechanistic models. Using transcranial magnetic stimulation as an example, we discuss how integrating methods with a focus on mechanisms could improve our predictions TMS effects within individuals, refine our models of health and disease, and improve our treatments.

Functional Connectivity within and beyond the Face Network Is Related to Reduced Discrimination of Degraded Faces in Young and Older Adults

Degrading face stimuli reduces face discrimination in both young and older adults, but the brain correlates of this decline in performance are not fully understood. We used functional magnetic resonance imaging to examine the effects of degraded face stimuli on face and nonface brain networks and tested whether these changes would predict the linear declines seen in performance. We found decreased activity in the face network (FN) and a decrease in the similarity of functional connectivity (FC) in the FN across conditions as degradation increased but no effect of age. FC in whole-brain networks also changed with increasing degradation, including increasing FC between the visual network and cognitive control networks. Older adults showed reduced modulation of this whole-brain FC pattern. The strongest predictors of within-participant decline in accuracy were changes in whole-brain network FC and FC similarity of the FN. There was no influence of age on these brain-behavior relations. These results suggest that a systems-level approach beyond the FN is required to understand the brain correlates of performance decline when faces are obscured with noise. In addition, the association between brain and behavior changes was maintained into older age, despite the dampened FC response to face degradation seen in older adults.

Individual Resting-State Brain Networks Enabled by Massive Multivariate Conditional Mutual Information

Individual-level resting-state networks (RSNs) based on resting-state fMRI (rs-fMRI) are of great interest due to evidence that network dysfunction may underlie some diseases. Most current rs-fMRI analyses use linear correlation. Since correlation is a bivariate measure of association, it discards most of the information contained in the spatial variation of the thousands of hemodynamic signals within the voxels in a given brain region. Subject-specific functional RSNs using typical rs-fMRI data, are therefore dominated by indirect connections and loss of spatial information and can only deliver reliable connectivity after group averaging. While bivariate partial correlation can rule out indirect connections, it results in connectivity that is too sparse due to lack of sensitivity. We have developed a method that uses all the spatial variation information in a given parcel by employing a multivariate information-theoretic association measure based on canonical correlations. Our method, multivariate conditional mutual information (mvCMI) reliably constructs single-subject connectivity estimates showing mostly direct connections. Averaging across subjects is not needed. The method is applied to Human Connectome Project data and compared to diffusion MRI. The results are far superior to those obtained by correlation and partial correlation.

An analytical workflow for seed-based correlation and independent component analysis in interventional resting-state fMRI studies

Resting-state functional MRI (rs-fMRI) is a task-free method of detecting spatially distinct brain regions with correlated activity, which form organised networks known as resting-state networks (RSNs). The two most widely used methods for analysing RSN connectivity are seed-based correlation analysis (SCA) and independent component analysis (ICA) but there is no established workflow of the optimal combination of analytical steps and how to execute them. Rodent rs-fMRI data from our previous longitudinal brain stimulation studies were used to investigate these two methods using FSL. Specifically, we examined: (1) RSN identification and group comparisons in ICA, (2) ICA-based denoising compared to nuisance signal regression in SCA, and (3) seed selection in SCA. In ICA, using a baseline-only template resulted in greater functional connectivity within RSNs and more sensitive detection of group differences than when an average pre/post stimulation template was used. In SCA, the use of an ICA-based denoising method in the preprocessing of rs-fMRI data and the use of seeds from individual functional connectivity maps in running group comparisons increased the sensitivity of detecting group differences by preventing the reduction in signals of interest. Accordingly, when analysing animal and human rs-fMRI data, we infer that the use of baseline-only templates in ICA and ICA-based denoising and individualised seeds in SCA will improve the reliability of results and comparability across rs-fMRI studies.

Impaired functional connectivity of sensorimotor network predicts recovery in drug-induced parkinsonism

OBJECTIVE

In a substantial portion of patients with drug-induced parkinsonism (DIP), parkinsonism may persist for long periods after discontinuation of offending drugs, suggesting subtle underlying neurodegeneration. We hypothesized that patients with DIP have impaired functional connectivity (FC) of brain networks, which may determine the reversibility of parkinsonism.

METHODS

In this case-control study, we consecutively recruited 60 patients with DIP and 32 healthy controls. We used independent component analysis and dual regression of functional magnetic resonance imaging data to identify seven resting-state networks and compared FC of the networks between the DIP and control groups. Among regions where the two groups showed a significant difference in the FC with sensorimotor network, we compared the FC between patients who had completely recovered (n = 21) and those who had partially recovered (n = 39) within 3 months of cessation of the offending drugs.

RESULTS

Patients with DIP had decreased FC between the sensorimotor network and widespread brain regions, when compared to healthy controls. FC in the prefrontal regions was negatively correlated with parkinsonian motor score. Patients who partially recovered had a significantly lower FC in the prefrontal and cerebellar regions than those who recovered completely, providing a useful predictor of recovery status.

CONCLUSIONS

Patients with DIP had decreased FC of the sensorimotor network, which correlated with the severity of parkinsonism and predicted the recovery status after cessation of offending drugs. Impaired FC of the sensorimotor network can be used as a biomarker to evaluate the severity and prognosis of DIP.

The effect of amyloid deposition on longitudinal resting-state functional connectivity in cognitively normal older adults

BACKGROUND

Pathological processes contributing to Alzheimer's disease begin decades prior to the onset of clinical symptoms. There is significant variation in cognitive changes in the presence of pathology, functional connectivity may be a marker of compensation to amyloid; however, this is not well understood.

METHODS

We recruited 64 cognitively normal older adults who underwent neuropsychological testing and biannual magnetic resonance imaging (MRI), amyloid imaging with Pittsburgh compound B (PiB)-PET, and glucose metabolism (FDG)-PET imaging for up to 6 years. Resting-state MRI was used to estimate connectivity of seven canonical neural networks using template-based rotation. Using voxel-wise paired t-tests, we identified neural networks that displayed significant changes in connectivity across time. We investigated associations among amyloid and longitudinal changes in connectivity and cognitive function by domains.

RESULTS

Left middle frontal gyrus connectivity within the memory encoding network increased over time, but the rate of change was lower with greater amyloid. This was no longer significant in an analysis where we limited the sample to only those with two time points. We found limited decline in cognitive domains overall. Greater functional connectivity was associated with better attention/processing speed and executive function (independent of time) in those with lower amyloid but was associated with worse function with greater amyloid.

CONCLUSIONS

Increased functional connectivity serves to preserve cognitive function in normal aging and may fail in the presence of pathology consistent with compensatory models.

Longitudinal degradation of the default/salience network axis in symptomatic individuals with elevated amyloid burden

Resting-state functional connectivity MRI (rs-fcMRI) is a non-invasive imaging technique that has come into increasing use to understand disrupted neural network function in neuropsychiatric disease. However, despite extensive study over the past 15 years, the development of rs-fcMRI as a biomarker has been impeded by a lack of reliable longitudinal rs-fcMRI measures. Here we focus on longitudinal change along the Alzheimer's disease (AD) trajectory and demonstrate the utility of Template Based Rotation (TBR) in detecting differential longitudinal rs-fcMRI change between higher and lower amyloid burden individuals with mildly impaired cognition. Specifically, we examine a small (N = 24), but densely sampled (~5 observations over ~3 years), cohort of symptomatic individuals with serial rs-fcMRI imaging and PiB-PET imaging for β-amyloid pathology. We observed longitudinal decline of the Default Mode and Salience network axis (DMN/SAL) among impaired individuals with high amyloid burden. No other networks showed differential change in high vs. low amyloid individuals over time. The standardized effect size of AD related DMN/SAL change is comparable to the standardized effect size of amyloid-related change on the mini-mental state exam (MMSE) and hippocampal volume (HV). Last, we show that the AD-related change in DMN/SAL connectivity is almost completely independent of change on MMSE or HV, suggesting that rs-fcMRI is sensitive to an aspect of AD progression that is not captured by these other measures. Together these analyses demonstrate that longitudinal rs-fcMRI using TBR can capture disease-relevant network disruption in a clinical population.

Anticholinergic Amnesia is Mediated by Alterations in Human Network Connectivity Architecture

Disrupted cholinergic neurotransmission plays a central role in Alzheimer's disease, medication-induced memory impairment, and delirium. At the systems level, this suggests anticholinergic drugs may alter the activity and interplay of anatomically distributed neural networks critical for memory function. Using a network-sensitive imaging technique (functional connectivity MRI) and a double-blind, crossover design, we examined the consequences of anticholinergic drug administration on episodic memory and functional network architecture in a group of clinically normal elderly. We observed that low-dose scopolamine (0.2 mg IV) decreased episodic memory performance and selectively decreased connectivity strength in 3 of 7 cortical networks. Both memory and connectivity effects were independent of β-amyloid burden. Drug-induced connectivity changes within the Default and Salience networks, as well as reductions in the strength of anticorrelation between these 2 networks, were sufficient to fully statistically mediate the effects of scopolamine on memory performance. These results provide experimental support for the importance of the Default and Salience networks to memory performance and suggest scopolamine-induced amnesia is underpinned by disrupted connectivity within and between these 2 networks. More broadly, these results support the potential utility of fcMRI as tool examine the systems-level pharmacology of psychoactive drugs.

Preferential degradation of cognitive networks differentiates Alzheimer's disease from ageing

Converging evidence from structural, metabolic and functional connectivity MRI suggests that neurodegenerative diseases, such as Alzheimer's disease, target specific neural networks. However, age-related network changes commonly co-occur with neuropathological cascades, limiting efforts to disentangle disease-specific alterations in network function from those associated with normal ageing. Here we elucidate the differential effects of ageing and Alzheimer's disease pathology through simultaneous analyses of two functional connectivity MRI datasets: (i) young participants harbouring highly-penetrant mutations leading to autosomal-dominant Alzheimer's disease from the Dominantly Inherited Alzheimer's Network (DIAN), an Alzheimer's disease cohort in which age-related comorbidities are minimal and likelihood of progression along an Alzheimer's disease trajectory is extremely high; and (ii) young and elderly participants from the Harvard Aging Brain Study, a cohort in which imaging biomarkers of amyloid burden and neurodegeneration can be used to disambiguate ageing alone from preclinical Alzheimer's disease. Consonant with prior reports, we observed the preferential degradation of cognitive (especially the default and dorsal attention networks) over motor and sensory networks in early autosomal-dominant Alzheimer's disease, and found that this distinctive degradation pattern was magnified in more advanced stages of disease. Importantly, a nascent form of the pattern observed across the autosomal-dominant Alzheimer's disease spectrum was also detectable in clinically normal elderly with clear biomarker evidence of Alzheimer's disease pathology (preclinical Alzheimer's disease). At the more granular level of individual connections between node pairs, we observed that connections within cognitive networks were preferentially targeted in Alzheimer's disease (with between network connections relatively spared), and that connections between positively coupled nodes (correlations) were preferentially degraded as compared to connections between negatively coupled nodes (anti-correlations). In contrast, ageing in the absence of Alzheimer's disease biomarkers was characterized by a far less network-specific degradation across cognitive and sensory networks, of between- and within-network connections, and of connections between positively and negatively coupled nodes. We go on to demonstrate that formalizing the differential patterns of network degradation in ageing and Alzheimer's disease may have the practical benefit of yielding connectivity measurements that highlight early Alzheimer's disease-related connectivity changes over those due to age-related processes. Together, the contrasting patterns of connectivity in Alzheimer's disease and ageing add to prior work arguing against Alzheimer's disease as a form of accelerated ageing, and suggest multi-network composite functional connectivity MRI metrics may be useful in the detection of early Alzheimer's disease-specific alterations co-occurring with age-related connectivity changes. More broadly, our findings are consistent with a specific pattern of network degradation associated with the spreading of Alzheimer's disease pathology within targeted neural networks.

Group ICA for identifying biomarkers in schizophrenia: 'Adaptive' networks via spatially constrained ICA show more sensitivity to group differences than spatio-temporal regression

Brain functional networks identified from fMRI data can provide potential biomarkers for brain disorders. Group independent component analysis (GICA) is popular for extracting brain functional networks from multiple subjects. In GICA, different strategies exist for reconstructing subject-specific networks from the group-level networks. However, it is unknown whether these strategies have different sensitivities to group differences and abilities in distinguishing patients. Among GICA, spatio-temporal regression (STR) and spatially constrained ICA approaches such as group information guided ICA (GIG-ICA) can be used to propagate components (indicating networks) to a new subject that is not included in the original subjects. In this study, based on the same a priori network maps, we reconstructed subject-specific networks using these two methods separately from resting-state fMRI data of 151 schizophrenia patients (SZs) and 163 healthy controls (HCs). We investigated group differences in the estimated functional networks and the functional network connectivity (FNC) obtained by each method. The networks were also used as features in a cross-validated support vector machine (SVM) for classifying SZs and HCs. We selected features using different strategies to provide a comprehensive comparison between the two methods. GIG-ICA generally showed greater sensitivity in statistical analysis and better classification performance (accuracy 76.45 ± 8.9%, sensitivity 0.74 ± 0.11, specificity 0.79 ± 0.11) than STR (accuracy 67.45 ± 8.13%, sensitivity 0.65 ± 0.11, specificity 0.71 ± 0.11). Importantly, results were also consistent when applied to an independent dataset including 82 HCs and 82 SZs. Our work suggests that the functional networks estimated by GIG-ICA are more sensitive to group differences, and GIG-ICA is promising for identifying image-derived biomarkers of brain disease.

•

We investigate which group ICA method is more sensitive to group differences in networks and FNC.

•

We evaluate which group ICA method can result in better classification performance.

•

We compare the spatio-temporal regression (STR) and GIG-ICA using different feature selection and model building strategies and independent samples.

•

We use different feature selection and model building strategies and classification using independent samples.

•

Using fMRI data including schizophrenia patients, GIG-ICA shows better performance than the traditional STR method.

Reprint of: Mapping human brain lesions and their functional consequences

Neuroscience has a long history of inferring brain function by examining the relationship between brain injury and subsequent behavioral impairments. The primary advantage of this method over correlative methods is that it can tell us if a certain brain region is necessary for a given cognitive function. In addition, lesion-based analyses provide unique insights into clinical deficits. In the last decade, statistical voxel-based lesion behavior mapping (VLBM) emerged as a powerful method for understanding the architecture of the human brain. This review illustrates how VLBM improves our knowledge of functional brain architecture, as well as how it is inherently limited by its mass-univariate approach. A wide array of recently developed methods appear to supplement traditional VLBM. This paper provides an overview of these new methods, including the use of specialized imaging modalities, the combination of structural imaging with normative connectome data, as well as multivariate analyses of structural imaging data. We see these new methods as complementing rather than replacing traditional VLBM, providing synergistic tools to answer related questions. Finally, we discuss the potential for these methods to become established in cognitive neuroscience and in clinical applications.

Mapping human brain lesions and their functional consequences

Neuroscience has a long history of inferring brain function by examining the relationship between brain injury and subsequent behavioral impairments. The primary advantage of this method over correlative methods is that it can tell us if a certain brain region is necessary for a given cognitive function. In addition, lesion-based analyses provide unique insights into clinical deficits. In the last decade, statistical voxel-based lesion behavior mapping (VLBM) emerged as a powerful method for understanding the architecture of the human brain. This review illustrates how VLBM improves our knowledge of functional brain architecture, as well as how it is inherently limited by its mass-univariate approach. A wide array of recently developed methods appear to supplement traditional VLBM. This paper provides an overview of these new methods, including the use of specialized imaging modalities, the combination of structural imaging with normative connectome data, as well as multivariate analyses of structural imaging data. We see these new methods as complementing rather than replacing traditional VLBM, providing synergistic tools to answer related questions. Finally, we discuss the potential for these methods to become established in cognitive neuroscience and in clinical applications.

Rapid whole-brain resting-state fMRI at 3 T: Efficiency-optimized three-dimensional EPI versus repetition time-matched simultaneous-multi-slice EPI

State-of-the-art simultaneous-multi-slice (SMS-)EPI and 3D-EPI share several properties that benefit functional MRI acquisition. Both sequences employ equivalent parallel imaging undersampling with controlled aliasing to achieve high temporal sampling rates. As a volumetric imaging sequence, 3D-EPI offers additional means of acceleration complementary to 2D-CAIPIRINHA sampling, such as fast water excitation and elliptical sampling. We performed an application-oriented comparison between a tailored, six-fold CAIPIRINHA-accelerated 3D-EPI protocol at 530 ms temporal and 2.4 mm isotropic spatial resolution and an SMS-EPI protocol with identical spatial and temporal resolution for whole-brain resting-state fMRI at 3 T. The latter required eight-fold slice acceleration to compensate for the lack of elliptical sampling and fast water excitation. Both sequences used vendor-supplied on-line image reconstruction. We acquired test/retest resting-state fMRI scans in ten volunteers, with simultaneous acquisition of cardiac and respiration data, subsequently used for optional physiological noise removal (nuisance regression). We found that the 3D-EPI protocol has significantly increased temporal signal-to-noise ratio throughout the brain as compared to the SMS-EPI protocol, especially when employing motion and nuisance regression. Both sequence types reliably identified known functional networks with stronger functional connectivity values for the 3D-EPI protocol. We conclude that the more time-efficient 3D-EPI primarily benefits from reduced parallel imaging noise due to a higher, actual k-space sampling density compared to SMS-EPI. The resultant BOLD sensitivity increase makes 3D-EPI a valuable alternative to SMS-EPI for whole-brain fMRI at 3 T, with voxel sizes well below 3 mm isotropic and sampling rates high enough to separate dominant cardiac signals from BOLD signals in the frequency domain.

Age-Related Increases in Tip-of-the-tongue are Distinct from Decreases in Remembering Names: A Functional MRI Study

Tip-of-the-tongue (TOT) experiences increase with age and frequently heighten concerns about memory decline. We studied 73 clinically normal older adults participating in the Harvard Aging Brain Study. They completed a functional magnetic resonance imaging (fMRI) task that required remembering names associated with pictures of famous faces. Older age was associated with more self-reported TOT experiences and a decrease in the percentage of remembered names. However, the percentage of TOT experiences and the percentage of remembered names were not directly correlated. We mapped fMRI activity for recollection of famous names and TOT and examined activity in the hippocampal formation, retrosplenial cortex, and lateral prefrontal cortex. The hippocampal formation was similarly activated in recollection and TOT experiences. In contrast, the retrosplenial cortex was most active for recollection and lateral prefrontal cortex was most active for TOT experiences. Together, the results confirm that age-related increases in TOT experiences are not only solely the consequence of age-related decline in recollection, but also likely reflect functional alterations in the brain networks that support retrieval monitoring and cognitive control. These findings provide behavioral and neuroimaging evidence that age-related TOT experiences and memory failure are partially independent processes.

Functional network integrity presages cognitive decline in preclinical Alzheimer disease

OBJECTIVE

To examine the utility of resting-state functional connectivity MRI (rs-fcMRI) measurements of network integrity as a predictor of future cognitive decline in preclinical Alzheimer disease (AD).

METHODS

A total of 237 clinically normal older adults (aged 63-90 years, Clinical Dementia Rating 0) underwent baseline β-amyloid (Aβ) imaging with Pittsburgh compound B PET and structural and rs-fcMRI. We identified 7 networks for analysis, including 4 cognitive networks (default, salience, dorsal attention, and frontoparietal control) and 3 noncognitive networks (primary visual, extrastriate visual, motor). Using linear and curvilinear mixed models, we used baseline connectivity in these networks to predict longitudinal changes in preclinical Alzheimer cognitive composite (PACC) performance, both alone and interacting with Aβ burden. Median neuropsychological follow-up was 3 years.

RESULTS

Baseline connectivity in the default, salience, and control networks predicted longitudinal PACC decline, unlike connectivity in the dorsal attention and all noncognitive networks. Default, salience, and control network connectivity was also synergistic with Aβ burden in predicting decline, with combined higher Aβ and lower connectivity predicting the steepest curvilinear decline in PACC performance.

CONCLUSIONS

In clinically normal older adults, lower functional connectivity predicted more rapid decline in PACC scores over time, particularly when coupled with increased Aβ burden. Among examined networks, default, salience, and control networks were the strongest predictors of rate of change in PACC scores, with the inflection point of greatest decline beyond the fourth year of follow-up. These results suggest that rs-fcMRI may be a useful predictor of early, AD-related cognitive decline in clinical research settings.

Phases of Hyperconnectivity and Hypoconnectivity in the Default Mode and Salience Networks Track with Amyloid and Tau in Clinically Normal Individuals

Alzheimer's disease (AD) is characterized by two hallmark molecular pathologies: amyloid aβ_1-42 and Tau neurofibrillary tangles. To date, studies of functional connectivity MRI (fcMRI) in individuals with preclinical AD have relied on associations with in vivo measures of amyloid pathology. With the recent advent of in vivo Tau-PET tracers it is now possible to extend investigations on fcMRI in a sample of cognitively normal elderly humans to regional measures of Tau. We modeled fcMRI measures across four major cortical association networks [default-mode network (DMN), salience network (SAL), dorsal attention network, and frontoparietal control network] as a function of global cortical amyloid [Pittsburgh Compound B (PiB)-PET] and regional Tau (AV1451-PET) in entorhinal, inferior temporal (IT), and inferior parietal cortex. Results showed that the interaction term between PiB and IT AV1451 was significantly associated with connectivity in the DMN and salience. The interaction revealed that amyloid-positive (aβ⁺) individuals show increased connectivity in the DMN and salience when neocortical Tau levels are low, whereas aβ⁺ individuals demonstrate decreased connectivity in these networks as a function of elevated Tau-PET signal. This pattern suggests a hyperconnectivity phase followed by a hypoconnectivity phase in the course of preclinical AD.SIGNIFICANCE STATEMENT This article offers a first look at the relationship between Tau-PET imaging with F¹⁸-AV1451 and functional connectivity MRI (fcMRI) in the context of amyloid-PET imaging. The results suggest a nonlinear relationship between fcMRI and both Tau-PET and amyloid-PET imaging. The pattern supports recent conjecture that the AD fcMRI trajectory is characterized by periods of both hyperconnectivity and hypoconnectivity. Furthermore, this nonlinear pattern can account for the sometimes conflicting reports of associations between amyloid and fcMRI in individuals with preclinical Alzheimer's disease.

Biomarkers, designs, and interpretations of resting-state fMRI in translational pharmacological research: A review of state-of-the-Art, challenges, and opportunities for studying brain chemistry

A decade of research and development in resting-state functional MRI (RSfMRI) has opened new translational and clinical research frontiers. This review aims to bridge between technical and clinical researchers who seek reliable neuroimaging biomarkers for studying drug interactions with the brain. About 85 pharma-RSfMRI studies using BOLD signal (75% of all) or arterial spin labeling (ASL) were surveyed to investigate the acute effects of psychoactive drugs. Experimental designs and objectives include drug fingerprinting dose-response evaluation, biomarker validation and calibration, and translational studies. Common biomarkers in these studies include functional connectivity, graph metrics, cerebral blood flow and the amplitude and spectrum of BOLD fluctuations. Overall, RSfMRI-derived biomarkers seem to be sensitive to spatiotemporal dynamics of drug interactions with the brain. However, drugs cause both central and peripheral effects, thus exacerbate difficulties related to biological confounds, structured noise from motion and physiological confounds, as well as modeling and inference testing. Currently, these issues are not well explored, and heterogeneities in experimental design, data acquisition and preprocessing make comparative or meta-analysis of existing reports impossible. A unifying collaborative framework for data-sharing and data-mining is thus necessary for investigating the commonalities and differences in biomarker sensitivity and specificity, and establishing guidelines. Multimodal datasets including sham-placebo or active control sessions and repeated measurements of various psychometric, physiological, metabolic and neuroimaging phenotypes are essential for pharmacokinetic/pharmacodynamic modeling and interpretation of the findings. We provide a list of basic minimum and advanced options that can be considered in design and analyses of future pharma-RSfMRI studies. Hum Brain Mapp 38:2276-2325, 2017. © 2017 Wiley Periodicals, Inc.

Harvard Aging Brain Study: Dataset and accessibility

The Harvard Aging Brain Study is sharing its data with the global research community. The longitudinal dataset consists of a 284-subject cohort with the following modalities acquired: demographics, clinical assessment, comprehensive neuropsychological testing, clinical biomarkers, and neuroimaging. To promote more extensive analyses, imaging data was designed to be compatible with other publicly available datasets. A cloud-based system enables access to interested researchers with blinded data available contingent upon completion of a data usage agreement and administrative approval. Data collection is ongoing and currently in its fifth year.

sGraSP: A graph-based method for the derivation of subject-specific functional parcellations of the brain

BACKGROUND

Resting-state fMRI (rs-fMRI) has emerged as a prominent tool for the study of functional connectivity. The identification of the regions associated with the different brain functions has received significant interest. However, most of the studies conducted so far have focused on the definition of a common set of regions, valid for an entire population. The variation of the functional regions within a population has rarely been accounted for.

NEW METHOD

In this paper, we propose sGraSP, a graph-based approach for the derivation of subject-specific functional parcellations. Our method generates first a common parcellation for an entire population, which is then adapted to each subject individually.

RESULTS

Several cortical parcellations were generated for 859 children being part of the Philadelphia Neurodevelopmental Cohort. The stability of the parcellations generated by sGraSP was tested by mixing population and subject rs-fMRI signals, to generate subject-specific parcels increasingly closer to the population parcellation. We also checked if the parcels generated by our method were better capturing a development trend underlying our data than the original parcels, defined for the entire population.

COMPARISON WITH EXISTING METHODS

We compared sGraSP with a simpler and faster approach based on a Voronoi tessellation, by measuring their ability to produce functionally coherent parcels adapted to the subject data.

CONCLUSIONS

Our parcellations outperformed the Voronoi tessellations. The parcels generated by sGraSP vary consistently with respect to signal mixing, the results are highly reproducible and the neurodevelopmental trend is better captured with the subject-specific parcellation, under all the signal mixing conditions.

Quantifying functional connectivity in multi-subject fMRI data using component models

Functional magnetic resonance imaging (fMRI) is increasingly used to characterize functional connectivity between brain regions. Given the vast number of between-voxel interactions in high-dimensional fMRI data, it is an ongoing challenge to detect stable and generalizable functional connectivity in the brain among groups of subjects. Component models can be used to define subspace representations of functional connectivity that are more interpretable. It is, however, unclear which component model provides the optimal representation of functional networks for multi-subject fMRI datasets. A flexible cross-validation approach that assesses the ability of the models to predict voxel-wise covariance in new data, using three different measures of generalization was proposed. This framework is used to compare a range of component models with varying degrees of flexibility in their representation of functional connectivity, evaluated on both simulated and experimental resting-state fMRI data. It was demonstrated that highly flexible subject-specific component subspaces, as well as very constrained average models, are poor predictors of whole-brain functional connectivity, whereas the best-generalizing models account for subject variability within a common spatial subspace. Within this set of models, spatial Independent Component Analysis (sICA) on concatenated data provides more interpretable brain patterns, whereas a consistent-covariance model that accounts for subject-specific network scaling (PARAFAC2) provides greater stability in functional connectivity relationships between components and their spatial representations. The proposed evaluation framework is a promising quantitative approach to evaluating component models, and reveals important differences between subspace models in terms of predictability, robustness, characterization of subject variability, and interpretability of the model parameters. Hum Brain Mapp 38:882-899, 2017. © 2016 Wiley Periodicals, Inc.

Resting-state fMRI study of acute migraine treatment with kinetic oscillation stimulation in nasal cavity

Kinetic oscillatory stimulation (KOS) in the nasal cavity is a non-invasive cranial nerve stimulation method with promising efficacy for acute migraine and other inflammatory disorders. For a better understanding of the underlying neurophysiological mechanisms of KOS treatment, we conducted a resting-state functional magnetic resonance imaging (fMRI) study of 10 acute migraine patients and 10 normal control subjects during KOS treatment in a 3 T clinical MRI scanner. The fMRI data were first processed using a group independent component analysis (ICA) method and then further analyzed with a voxel-wise 3-way ANOVA modeling and region of interest (ROI) of functional connectivity metrics.

All migraine participants were relieved from their acute migraine symptoms after 10–20 min KOS treatment and remained migraine free for 3–6 months. The resting-state fMRI result indicates that migraine patients have altered intrinsic functional activity in the anterior cingulate, inferior frontal gyrus and middle/superior temporal gyrus. KOS treatment gave rise to up-regulated intrinsic functional activity for migraine patients in a number of brain regions involving the limbic and primary sensory systems, while down regulating temporally the activity for normal controls in a few brain areas, such as the right dorsal posterior insula and inferior frontal gyrus.

The result of this study confirms the efficacy of KOS treatment for relieving acute migraine symptoms and reducing attack frequency. Resting-state fMRI measurements demonstrate that migraine is associated with aberrant intrinsic functional activity in the limbic and primary sensory systems. KOS in the nasal cavity gives rise to the adjustment of the intrinsic functional activity in the limbic and primary sensory networks and restores the physiological homeostasis in the autonomic nervous system.

•

Efficacy and neurological mechanisms underlying kinetic oscillatory stimulation treatment of migraine

•

Dependence of ICA (independent component analysis) results on the number of independent components.

•

Modulation of ANS (autonomic nervous system) function via the limbic network

Altered intrinsic organisation of brain networks implicated in attentional processes in adult attention-deficit/hyperactivity disorder: a resting-state study of attention, default mode and salience network connectivity

Deficits in task-related attentional engagement in attention-deficit/hyperactivity disorder (ADHD) have been hypothesised to be due to altered interrelationships between attention, default mode and salience networks. We examined the intrinsic connectivity during rest within and between these networks. Six-minute resting-state scans were obtained. Using a network-based approach, connectivity within and between the dorsal and ventral attention, the default mode and the salience networks was compared between the ADHD and control group. The ADHD group displayed hyperconnectivity between the two attention networks and within the default mode and ventral attention network. The salience network was hypoconnected to the dorsal attention network. There were trends towards hyperconnectivity within the dorsal attention network and between the salience and ventral attention network in ADHD. Connectivity within and between other networks was unrelated to ADHD. Our findings highlight the altered connectivity within and between attention networks, and between them and the salience network in ADHD. One hypothesis to be tested in future studies is that individuals with ADHD are affected by an imbalance between ventral and dorsal attention systems with the former playing a dominant role during task engagement, making individuals with ADHD highly susceptible to distraction by salient task-irrelevant stimuli.

Artifact removal in the context of group ICA: A comparison of single-subject and group approaches

Independent component analysis (ICA) has been widely applied to identify intrinsic brain networks from fMRI data. Group ICA computes group-level components from all data and subsequently estimates individual-level components to recapture intersubject variability. However, the best approach to handle artifacts, which may vary widely among subjects, is not yet clear. In this work, we study and compare two ICA approaches for artifacts removal. One approach, recommended in recent work by the Human Connectome Project, first performs ICA on individual subject data to remove artifacts, and then applies a group ICA on the cleaned data from all subjects. We refer to this approach as Individual ICA based artifacts Removal Plus Group ICA (IRPG). A second proposed approach, called Group Information Guided ICA (GIG-ICA), performs ICA on group data, then removes the group-level artifact components, and finally performs subject-specific ICAs using the group-level non-artifact components as spatial references. We used simulations to evaluate the two approaches with respect to the effects of data quality, data quantity, variable number of sources among subjects, and spatially unique artifacts. Resting-state test-retest datasets were also employed to investigate the reliability of functional networks. Results from simulations demonstrate GIG-ICA has greater performance compared with IRPG, even in the case when single-subject artifacts removal is perfect and when individual subjects have spatially unique artifacts. Experiments using test-retest data suggest that GIG-ICA provides more reliable functional networks. Based on high estimation accuracy, ease of implementation, and high reliability of functional networks, we find GIG-ICA to be a promising approach.

Functional Connectivity in Multiple Cortical Networks Is Associated with Performance Across Cognitive Domains in Older Adults

Intrinsic functional connectivity MRI has become a widely used tool for measuring integrity in large-scale cortical networks. This study examined multiple cortical networks using Template-Based Rotation (TBR), a method that applies a priori network and nuisance component templates defined from an independent dataset to test datasets of interest. A priori templates were applied to a test dataset of 276 older adults (ages 65-90) from the Harvard Aging Brain Study to examine the relationship between multiple large-scale cortical networks and cognition. Factor scores derived from neuropsychological tests represented processing speed, executive function, and episodic memory. Resting-state BOLD data were acquired in two 6-min acquisitions on a 3-Tesla scanner and processed with TBR to extract individual-level metrics of network connectivity in multiple cortical networks. All results controlled for data quality metrics, including motion. Connectivity in multiple large-scale cortical networks was positively related to all cognitive domains, with a composite measure of general connectivity positively associated with general cognitive performance. Controlling for the correlations between networks, the frontoparietal control network (FPCN) and executive function demonstrated the only significant association, suggesting specificity in this relationship. Further analyses found that the FPCN mediated the relationships of the other networks with cognition, suggesting that this network may play a central role in understanding individual variation in cognition during aging.

Challenges in the reproducibility of clinical studies with resting state fMRI: An example in early Parkinson's disease

Resting state fMRI (rfMRI) is gaining in popularity, being easy to acquire and with promising clinical applications. However, rfMRI studies, especially those involving clinical groups, still lack reproducibility, largely due to the different analysis settings. This is particularly important for the development of imaging biomarkers. The aim of this work was to evaluate the reproducibility of our recent study regarding the functional connectivity of the basal ganglia network in early Parkinson's disease (PD) (Szewczyk-Krolikowski et al., 2014). In particular, we systematically analysed the influence of two rfMRI analysis steps on the results: the individual cleaning (artefact removal) of fMRI data and the choice of the set of independent components (template) used for dual regression.

Our experience suggests that the use of a cleaning approach based on single-subject independent component analysis, which removes non neural-related sources of inter-individual variability, can help to increase the reproducibility of clinical findings. A template generated using an independent set of healthy controls is recommended for studies where the aim is to detect differences from a “healthy” brain, rather than an “average” template, derived from an equal number of patients and controls. While, exploratory analyses (e.g. testing multiple resting state networks) should be used to formulate new hypotheses, careful validation is necessary before promising findings can be translated into useful biomarkers.

•

Reproducibility of clinical findings is crucial for imaging biomarker development.

•

We addressed the impact on reproducibility of different analysis settings in rfMRI.

•

ICA-based cleaning of rfMRI data increases reproducibility.

•

The effect of the template choice for dual regression is evaluated.

Network localization of neurological symptoms from focal brain lesions

A traditional and widely used approach for linking neurological symptoms to specific brain regions involves identifying overlap in lesion location across patients with similar symptoms, termed lesion mapping. This approach is powerful and broadly applicable, but has limitations when symptoms do not localize to a single region or stem from dysfunction in regions connected to the lesion site rather than the site itself. A newer approach sensitive to such network effects involves functional neuroimaging of patients, but this requires specialized brain scans beyond routine clinical data, making it less versatile and difficult to apply when symptoms are rare or transient. In this article we show that the traditional approach to lesion mapping can be expanded to incorporate network effects into symptom localization without the need for specialized neuroimaging of patients. Our approach involves three steps: (i) transferring the three-dimensional volume of a brain lesion onto a reference brain; (ii) assessing the intrinsic functional connectivity of the lesion volume with the rest of the brain using normative connectome data; and (iii) overlapping lesion-associated networks to identify regions common to a clinical syndrome. We first tested our approach in peduncular hallucinosis, a syndrome of visual hallucinations following subcortical lesions long hypothesized to be due to network effects on extrastriate visual cortex. While the lesions themselves were heterogeneously distributed with little overlap in lesion location, 22 of 23 lesions were negatively correlated with extrastriate visual cortex. This network overlap was specific compared to other subcortical lesions (P < 10(-5)) and relative to other cortical regions (P < 0.01). Next, we tested for generalizability of our technique by applying it to three additional lesion syndromes: central post-stroke pain, auditory hallucinosis, and subcortical aphasia. In each syndrome, heterogeneous lesions that themselves had little overlap showed significant network overlap in cortical areas previously implicated in symptom expression (P < 10(-4)). These results suggest that (i) heterogeneous lesions producing similar symptoms share functional connectivity to specific brain regions involved in symptom expression; and (ii) publically available human connectome data can be used to incorporate these network effects into traditional lesion mapping approaches. Because the current technique requires no specialized imaging of patients it may prove a versatile and broadly applicable approach for localizing neurological symptoms in the setting of brain lesions.

Putamen-midbrain functional connectivity is related to striatal dopamine transporter availability in patients with Lewy body diseases

Prior work has shown that functional connectivity between the midbrain and putamen is altered in patients with impairments in the dopamine system. This study examines whether individual differences in midbrain-striatal connectivity are proportional to the integrity of the dopamine system in patients with nigrostriatal dopamine loss (Parkinson's disease and dementia with Lewy bodies). We assessed functional connectivity of the putamen during resting state fMRI and dopamine transporter (DAT) availability in the striatum using 11C-Altropane PET in twenty patients. In line with the hypothesis that functional connectivity between the midbrain and the putamen reflects the integrity of the dopaminergic neurotransmitter system, putamen-midbrain functional connectivity was significantly correlated with striatal DAT availability even after stringent control for effects of head motion. DAT availability did not relate to functional connectivity between the caudate and thalamus/prefrontal areas. As such, resting state functional connectivity in the midbrain-striatal pathway may provide a useful indicator of underlying pathology in patients with nigrostriatal dopamine loss.

Decoupling of structural and functional brain connectivity in older adults with white matter hyperintensities

Age-related impairments in the default network (DN) have been related to disruptions in connecting white matter tracts. We hypothesized that the local correlation between DN structural and functional connectivity is negatively affected in the presence of global white matter injury. In 125 clinically normal older adults, we tested whether the relationship between structural connectivity (via diffusion imaging tractography) and functional connectivity (via resting-state functional MRI) of the posterior cingulate cortex (PCC) and medial prefrontal frontal cortex (MPFC) of the DN was altered in the presence of white matter hyperintensities (WMH). A significant correlation was observed between microstructural properties of the cingulum bundle and MPFC-PCC functional connectivity in individuals with low WMH load, but not with high WMH load. No correlation was observed between PCC-MPFC functional connectivity and microstructure of the inferior longitudinal fasciculus, a tract not passing through the PCC or MPFC. Decoupling of connectivity, measured as the absolute difference between structural and functional connectivity, in the high WMH group was related to poorer executive functioning and memory performance. These results suggest that such decoupling may reflect reorganization of functional networks in response to global white matter pathology and may provide an early marker of clinically relevant network alterations.

Dimensionality of ICA in resting-state fMRI investigated by feature optimized classification of independent components with SVM

Different machine learning algorithms have recently been used for assisting automated classification of independent component analysis (ICA) results from resting-state fMRI data. The success of this approach relies on identification of artifact components and meaningful functional networks. A limiting factor of ICA is the uncertainty of the number of independent components (NIC). We aim to develop a framework based on support vector machines (SVM) and optimized feature-selection for automated classification of independent components (ICs) and use the framework to investigate the effects of input NIC on the ICA results. Seven different resting-state fMRI datasets were studied. 18 features were devised by mimicking the empirical criteria for manual evaluation. The five most significant (p < 0.01) features were identified by general linear modeling and used to generate a classification model for the framework. This feature-optimized classification of ICs with SVM (FOCIS) framework was used to classify both group and single subject ICA results. The classification results obtained using FOCIS and previously published FSL-FIX were compared against manually evaluated results. On average the false negative rate in identifying artifact contaminated ICs for FOCIS and FSL-FIX were 98.27 and 92.34%, respectively. The number of artifact and functional network components increased almost linearly with the input NIC. Through tracking, we demonstrate that incrementing NIC affects most ICs when NIC < 33, whereas only a few limited ICs are affected by direct splitting when NIC is incremented beyond NIC > 40. For a given IC, its changes with increasing NIC are individually specific irrespective whether the component is a potential resting-state functional network or an artifact component. Using FOCIS, we investigated experimentally the ICA dimensionality of resting-state fMRI datasets and found that the input NIC can critically affect the ICA results of resting-state fMRI data.

Multiple Brain Markers are Linked to Age-Related Variation in Cognition

Age-related alterations in brain structure and function have been challenging to link to cognition due to potential overlapping influences of multiple neurobiological cascades. We examined multiple brain markers associated with age-related variation in cognition. Clinically normal older humans aged 65-90 from the Harvard Aging Brain Study (N = 186) were characterized on a priori magnetic resonance imaging markers of gray matter thickness and volume, white matter hyperintensities, fractional anisotropy (FA), resting-state functional connectivity, positron emission tomography markers of glucose metabolism and amyloid burden, and cognitive factors of processing speed, executive function, and episodic memory. Partial correlation and mediation analyses estimated age-related variance in cognition shared with individual brain markers and unique to each marker. The largest relationships linked FA and striatum volume to processing speed and executive function, and hippocampal volume to episodic memory. Of the age-related variance in cognition, 70-80% was accounted for by combining all brain markers (but only ∼20% of total variance). Age had significant indirect effects on cognition via brain markers, with significant markers varying across cognitive domains. These results suggest that most age-related variation in cognition is shared among multiple brain markers, but potential specificity between some brain markers and cognitive domains motivates additional study of age-related markers of neural health.