Fusing functional signals by sparse canonical correlation analysis improves network reproducibility

We contribute a novel multivariate strategy for computing the structure of functional networks in the brain from arterial spin labeling (ASL) MRI. Our method fuses and correlates multiple functional signals by employing an interpretable dimensionality reduction method, sparse canonical correlation analysis (SCCA). There are two key aspects of this contribution. First, we show how SCCA may be used to compute a multivariate correlation between different regions of interest (ROI). In contrast to averaging the signal over the ROI, this approach exploits the full information within the ROI. Second, we show how SCCA may simultaneously exploit both the ASL-BOLD and ASL-based cerebral blood flow (CBF) time series to produce network measurements. Our approach to fusing multiple time signals in network studies improves reproducibility over standard approaches while retaining the interpretability afforded by the classic ROI region-averaging methods. We show experimentally in test-retest data that our sparse CCA method extracts biologically plausible and stable functional network structures from ASL. We compare the ROI approach to the CCA approach while using CBF measurements alone. We then compare these results to the joint BOLD-CBF networks in a reproducibility study and in a study of functional network structure in traumatic brain injury (TBI). Our results show that the SCCA approach provides significantly more reproducible results compared to region-averaging, and in TBI the SCCA approach reveals connectivity differences not seen with the region averaging approach.

For each metric, using both region averaging (orange) and SCCA (yellow), connectivity matrices were calculated from ASL data acquired in separate acquisitions in the same day and for data acquired one week apart. Whole network correlations were then calculated to examine reliability for the daily (left) and weekly (right) data for each subject. Here we illustrate results using sparsity values of s=t=0.05. A range of sparsity values (s=t) up to 0.25 were examined and these higher values did not produce qualitatively different results.

For each metric, using both region averaging (orange) and SCCA (yellow), connectivity matrices were calculated from ASL data acquired in separate acquisitions in the same day and for data acquired one week apart. Whole network correlations were then calculated to examine reliability for the daily (left) and weekly (right) data for each subject. Here we illustrate results using sparsity values of s=t=0.05. A range of sparsity values (s=t) up to 0.25 were examined and these higher values did not produce qualitatively different results.

J. T. Duda, J. A. Detre, J. Kim, J. Gee, and B. B. Avants, “Fusing functional signals by sparse canonical correlation analysis improves network reproducibility,” in Medical image computing and computer-assisted intervention–miccai 2013, Springer, 2013, pp. 635-642.

Advertisements

Relating Cerebral Blood Flow to Structural & Functional Metrics in Typically Developing Children


Sample slices from the multivariate atlas used as a basis for neuro-anatomical comparison.

Purpose: To evaluate the relationships between cerebral blood flow and other magnetic resonance (MR) imaging based measures such as fractional anisotropy, magnetic transfer ratio, cortical thickness and mean resting state BOLD signal in typically developing children.

Methods: Eighty-eight children aged 7-17 underwent pseudo-continuous arterial spin-labeled perfusion MRI (pCASL) [1] examinations along with anatomical (T1), diffusion tensor (DTI), magnetic transfer (MT) and BOLD resting state functional MRI (rs-fMRI) examinations. For each imaging modality, the ANTs [2] toolkit was used to create a modality-specific template from a subset (n=30) of the subjects. For non-scalar modalities, derived scalar images were used for template building. For pCASL the mean CBF image was used; for DTI the average diffusion weighted image was used; for rs-fMRI the mean BOLD image was used; and for MT the M0 image was used. Each modality-specific template was then registered to the T1 template to obtain a single multi-modality template (MMT). The T1 component of the MMT was then brain-masked, labeled, and three-tissue segmented using the Atropos segmentation tool [2]. For each subject, each modality was aligned to the corresponding component of the MMT for brain-masking and labeling. Intra-subject registrations were then performed to align all modalities to each subject’s T1 image. To provide a basis for comparison, a scalar metric was derived for each image modality. For pCASL the mean CBF was calculated; for T1 images, the cortical thickness was measured using the DiRECT method; fractional anisotropy was calculated from the DTI; the magnetization transfer ratio (MTR) was calculated from the MT images; and mean BOLD signal was calculated from the resting state fMRI data.

Results: Regularized canonical correlation analysis, as implemented in the sscan tool [2], was used to identify the relationship between CBF and each of the additional modalities. The analysis of each modality type is restricted to the most informative tissue type for that modality. For CBF, rs-fMRI and cortical thickness, only values in gray matter are examined, while only values in white matter are examined for FA and MTR.

Discussion: To the best of our knowledge, this is the first study to simultaneously compare CBF to cortical thickness, fractional anisotropy, magnetization transfer ratio and mean resting BOLD signal in a single population. In doing so, we hope to gain insight regarding the degree to which CBF provides statistically unique information in relation to these additional MR imaging modalities. Additionally, the development of the framework for analyzing these modalities provides a basis for future studies to explore the relationship between CBF and network based measures of both structural and functional connectivity.

Conclusion: The relationship between cortical thickness and Mean CBF (R2=0.4777) was the strongest of the metrics examined. In white matter, the MTR (R2=0.3126)  was stronger than FA (R2=0.1462). The mean BOLD (R2=0.1414) metric was the weakest.

Poster PDF

Self link – http://goo.gl/40OiM

Structural and functional connectivity have network-wide influences upon cognitive performance

In this paper functional subnetworks in the brain were examined using MRI to measure both structural connectivity and functional connectivity. Additionally, the influence on behavior of both types of connectivity examined to determine the degree to which each provides unique information as well as how this information may be used to identify the parts of a network that are most influential on behavioral performance. Functional connectivity involves co-activation of brain regions during performance of a task while brain recruitment is monitored with fMRI. Structural connectivity is related to the long tract white matter projections that may integrate recruited brain regions biologically. Here we demonstrate how structural and functional connectivity may be used to examine small, functionally defined subnetworks in the brain during performance of a common language task. Functionally defined cortical regions are used along with a population-averaged diffusion tensor atlas to identify the white matter pathways that provide the basis for biological connectivity. A centerline-based method is used to provide a geometric model that facilitates the equidimensional comparison of functional and structural connectivity within a network. Behavioral data are used to identify the relative contributions of function and structure, and the degree to which each provides unique insight into behavior.

Duda, Jeffrey T., “Characterizing Connectivity In Brain Networks Using Magnetic Resonance Imaging” (2010). Publicly accessible Penn Dissertations. Paper 191.