Examining the spontaneous activity to understand the neural mechanism of brain disorders and establish
neuroimaging-based disease-related biomarkers is a focus in recent resting-state functional MRI (fMRI) studies. The
present study hypothesized that resting activity in the default mode network (DMN), which was used for characterizing
the resting-state human brain might be different in patients with depressed Parkinson disease (dPD) compared with
non-depressed Parkinson disease (ndPD) patients. To test the hypothesis, we firstly employed the Group independent
component analysis (ICA) approach to isolate the DMN for the two groups by analyzing the resting-state fMRI data from
a group of 12 patients with dPD and a group of 12 age-matched ndPD subjects. Between-group comparison of the
functional connectivity in the DMN was then performed to examine the impact of depression on the intrinsic activity in
PD. We found 1) the core region from the network the medial prefrontal cortex (MPFC) show significant decreased
activity in dPD group compared with ndPD group; 2) the activity in MPFC has significant negative correlation with
behavioral measure; 3) the resting activity intensity of MPFC is suggested to be a promising biomarker for distinguishing
dPD from ndPD.
Task-based neuroimaging studies revealed that different attention operations were carried out by the functional
interaction and cooperation between two attention systems: the dorsal attention network (DAN) and the ventral attention
network (VAN), which were respectively involved in the "top-down" endogenous attention orienting and the "bottomup"
exogenous attention reorienting process. Recent focused resting functional MRI (fMRI) studies found the two
attention systems were inherently organized in the human brain regardless of whether or not the attention process were
required, but how the two attention systems interact with each other in the absence of task is yet to be investigated. In
this study, we first separated the DAN and VAN by applying the group independent component analysis (ICA) to the
resting fMRI data acquired from 12 healthy young subjects, then used Gaussian Bayesian network (BN) learning
approach to explore the plausible effective connectivity pattern of the two attention systems. It was found regions from
the same attention network were strongly intra-dependent, and all the connections were located in the information flow
from VAN to DAN, which suggested that an orderly functional interactions and information exchanges between the two
attention networks existed in the intrinsic spontaneous brain activity, and the inherent connections might benefit the
efficient cognitive process between DAN and VAN, such as the "top-down" and "bottom-up" reciprocal interaction when
attention-related tasks were involved.
Spatial Independent component analysis (sICA) has been successfully used to analyze functional magnetic resonance
(fMRI) data. However, the application of ICA was limited in multi-task fMRI data due to the potential spatial
dependence between task-related components. Long et al. (2009) proposed ICA with linear projection (ICAp) method
and demonstrated its capacity to solve the interaction among task-related components in multi-task fMRI data of single
subject. However, it's unclear that how to perform ICAp over a group of subjects. In this study, we proposed a group
analysis framework on multi-task fMRI data by combining ICAp with the temporal concatenation method reported by
Calhoun (2001). The results of real fMRI experiment containing multiple visual processing tasks demonstrated the
feasibility and effectiveness of the group ICAp method. Moreover, compared to the GLM method, the group ICAp
method is more sensitive to detect the regions specific to each task.
KEYWORDS: Brain, Functional magnetic resonance imaging, Independent component analysis, Magnetic resonance imaging, Brain mapping, Neuroimaging, Scanners, Data processing, Data acquisition, Principal component analysis
Resting-state functional MRI (fMRI) studies have suggested the posterior cingulate cortex (PCC) plays a pivotal role in
the default mode network (DMN), a set of co-activated brain regions characterizing the resting-state brain. Concerning
this finding we propose the following questions in this study: Does PCC consistently play the equally crucial role in the
DMN across different subjects, such as healthy young and healthy old subjects? Whether the fMRI scan environments or
parameters would affect the results? To address these questions, we collected resting-state fMRI data on four groups of
subjects: two healthy young groups scanned under 3-T and 1.5-T MRI systems respectively, and two healthy elderly
groups both scanned under 3-T MRI system but with different scan parameters. Then group independent component
analysis was used to isolate the DMN, and partial correlation analysis was employed to reveal the direct interactions
between brain regions from the DMN. Finally, we measured the connectivity between brain regions based on the number
of significantly interacted links to every region within this network. We found that PCC was the brain region consistently
having the largest number of directly interacted regions in the four groups, suggesting the pivotal role of PCC in the
DMN was stable and consistent across healthy subjects. The results also suggested the function of PCC would be more
critical in healthy elderly subjects compared with healthy young subjects. And the factors of scan environments and
parameters did not show any obvious impact on the above conclusions in this investigation.
KEYWORDS: Functional magnetic resonance imaging, Independent component analysis, Data modeling, Brain, Data analysis, Neural networks, Magnetic resonance imaging, Data processing, Convolution, Feature extraction
The existence of the potential non-independency between task-related components in multi-task functional magnetic
resonance imaging (fMRI) studies limits the general application of Independent Component Analysis (ICA) method. The
ICA with projection (ICAp) method proposed by Long (2009, HBM) demonstrated its capacity to solve the interaction
among task-related components of multi-task fMRI data. The basic idea of projection is to remove the influence of the
uninteresting tasks through projection in order to extract one interesting task-related component. However, both the
stimulus paradigm of each task and the homodynamic response function (HRF) are essential for the projection. Due to
the noises in the data and the variability of the HRF across the voxels and subjects, the ideal time course of each task for
projection would be deviant from the true value, which might worsen the ICAp results. In order to make the time courses
for projection closer to the true value, the iterative ICAp is proposed in this study. The iterative ICAp is based on the
assumption that the task-related time courses extracted from the fMRI data by ICAp is more approximate to the true
value than the ideal reference function. Simulated experiment proved that both the spatial detection power and the
temporal accuracy of time course were increased for each task-related component. Moreover, the results of the real
two-task fMRI data were also improved by the iterative ICAp method.
KEYWORDS: Brain, Independent component analysis, Functional magnetic resonance imaging, Statistical analysis, Neuroimaging, Testing and analysis, Data modeling, Principal component analysis, Data acquisition, Analytical research
This work proposed to use the linear Gaussian Bayesian network (BN) to construct the effective connectivity model of
the brain's default mode network (DMN), a set of regions characterized by more increased neural activity during
rest-state than most goal-oriented tasks. In a complete unsupervised data-driven manner, Bayesian information criterion
(BIC) based learning approach was utilized to identify a highest scored network whose nodes (brain regions) were
selected based on the result from the group independent component analysis (Group ICA) examining the DMN. We put
forward to adopt the statistical significance testing method for regression coefficients used in stepwise regression
analysis to further refine the network identified by BIC. The final established BN, learned from the functional magnetic
resonance imaging (fMRI) data acquired from 12 healthy young subjects during rest-state, revealed that the hippocampus
(HC) was the most influential brain region that affected activities in all other regions included in the BN. In contrast, the
posterior cingulate cortex (PCC) was influenced by other regions, but had no reciprocal effects on any other region.
Overall, the configuration of our BN illustrated that a prominent connection from HC to PCC existed in the DMN.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.