Ranking fMRI time courses by minimum spanning trees: assessing coactivation in fMRI

Clustering the Brain With "CluB": A New Toolbox for Quantitative Meta-Analysis of Neuroimaging Data

In this paper we describe and validate a new coordinate-based method for meta-analysis of neuroimaging data based on an optimized hierarchical clustering algorithm: CluB (Clustering the Brain). The CluB toolbox permits both to extract a set of spatially coherent clusters of activations from a database of stereotactic coordinates, and to explore each single cluster of activation for its composition according to the cognitive dimensions of interest. This last step, called “cluster composition analysis,” permits to explore neurocognitive effects by adopting a factorial-design logic and by testing the working hypotheses using either asymptotic tests, or exact tests either in a classic inference, or in a Bayesian-like context. To perform our validation study, we selected the fMRI data from 24 normal controls involved in a reading task. We run a standard random-effects second level group analysis to obtain a “Gold Standard” of reference. In a second step, the subject-specific reading effects (i.e., the linear t-contrast “reading > baseline”) were extracted to obtain a coordinates-based database that was used to run a meta-analysis using both CluB and the popular Activation Likelihood Estimation method implemented in the software GingerALE. The results of the two meta-analyses were compared against the “Gold Standard” to compute performance measures, i.e., sensitivity, specificity, and accuracy. The GingerALE method obtained a high level of accuracy (0.967) associated with a high sensitivity (0.728) and specificity (0.971). The CluB method obtained a similar level of accuracy (0.956) and specificity (0.969), notwithstanding a lower level of sensitivity (0.14) due to the lack of prior Gaussian transformation of the data. Finally, the two methods obtained a good-level of concordance (AC₁ = 0.93). These results suggested that methods based on hierarchical clustering (and post-hoc statistics) and methods requiring prior Gaussian transformation of the data can be used as complementary tools, with the GingerALE method being optimal for neurofunctional mapping of pooled data according to simpler designs, and the CluB method being preferable to test more specific, and localized, neurocognitive hypotheses according to factorial designs.

The role of diversity in complex ICA algorithms for fMRI analysis

BACKGROUND

The widespread use of data-driven methods, such as independent component analysis (ICA), for the analysis of functional magnetic resonance imaging data (fMRI) has enabled deeper understanding of neural function. However, most popular ICA algorithms for fMRI analysis make several simplifying assumptions, thus ignoring sources of statistical information, types of "diversity," and limiting their performance.

NEW METHOD

We propose the use of complex entropy rate bound minimization (CERBM) for the analysis of actual fMRI data in its native, complex, domain. Though CERBM achieves enhanced performance through the exploitation of the three types of diversity inherent to complex fMRI data: noncircularity, non-Gaussianity, and sample-to-sample dependence, CERBM produces results that are more variable than simpler methods. This motivates the development of a minimum spanning tree (MST)-based stability analysis that mitigates the variability of CERBM.

COMPARISON WITH EXISTING METHODS

In order to validate our method, we compare the performance of CERBM with the popular CInfomax as well as complex entropy bound minimization (CEBM).

RESULTS

We show that by leveraging CERBM and the MST-based stability analysis, we are able to consistently produce components that have a greater number of activated voxels in physically meaningful regions and can more accurately classify patients with schizophrenia than components generated using simpler models.

CONCLUSIONS

Our results demonstrate the advantages of using ICA algorithms that can exploit all inherent types of diversity for the analysis of fMRI data when coupled with appropriate stability analyses.

A method for quantitative SEMG decomposition and MUAP classification during voluntary isovelocity elbow flexion

The purpose of this study was to develop an algorithm for surface electromyogram (SEMG) decomposition and classification of surface motor unit (MU) action potential (MUAP) detected during isovelocity elbow flexion. In our proposed algorithm, firstly the measured SEMG was extracted for 3 seconds by every 1.5 seconds. SEMG was decomposed with Independent Component Analysis (ICA) technique, and classified with template matching. Finally, the MUAP trains were identified under the firing time of the MUAPs classified in each extracted period. The SEMG was measured from the biceps short head muscle during voluntary elbow flexion of 0 to 90 degrees at constant velocity 9 degree/s against a constant load torque of 10%MVC and the MUAPs were classified with our proposed algorithm. As a result, calculated MUs firing rates were almost same as the results in the previous studies. It was shown that the proposed algorithm was useful for decomposing SEMG detected during flexion movements.

Neural networks approach to clustering of activity in fMRI data

Clusters of correlated activity in functional magnetic resonance imaging data can identify regions of interest and indicate interacting brain areas. Because the extraction of clusters is computationally complex, we apply an approximative method which is based on artificial neural networks. It allows one to find clusters of various degrees of connectivity ranging between the two extreme cases of cliques and connectivity components. We propose a criterion which allows to evaluate the relevance of such structures based on the robustness with respect to parameter variations. Exploiting the intracluster correlations, we can show that regions of substantial correlation with an external stimulus can be unambiguously separated from other activity.

Single-trial variability in early visual neuromagnetic responses: an explorative study based on the regional activation contributing to the N70m peak

Cortical activity evoked by repeated identical sensory stimulation is extremely variable. The source of this variability is often assigned to "random ongoing background activity" which is considered to be irrelevant to the processing of the stimuli and can therefore be eliminated by ensemble averaging. In this work, we studied the single-trial variability in neuromagnetic responses elicited by circular checkerboard pattern stimuli with radii of 1.8 degrees, 3.7 degrees, and 4.5 degrees. For most of the MEG sensors over the occipital areas, the averaged signal showed a clear early (N70m) response following the stimulus onset and this response was modulated by the checkerboard size. A data-driven spatial filter was used to extract one of the many possible composite time courses of single-trial activity corresponding to the complex of N70m generators. Pattern analysis principles were then employed to analyze, classify, and handle the extracted temporal patterns. We explored whether these patterns correspond to distinct response modes, which could characterize the evoked response better than the averaged signal and over an extended range of latencies around N70m. A novel scheme for detecting and organizing the structure in single-trial recordings was utilized. This served as a basis for comparisons between runs with different checkerboard sizes and provided a causal interpretation of variability in terms of regional dynamics, including the relatively weak activation in primary visual cortex. At the level of single trial activity, the polymorphic response to a simple stimulus is generated by a coupling of polymodal areas and cooperative activity in striate and extrastriate areas. Our results suggest a state-dependent response with a wide range of characteristic time scales and indicate the ongoing activity as a marker of the responsiveness state.

Multidimensional scaling of integrated neurocognitive function and schizophrenia as a disconnexion disorder

Multidimensional scaling (MDS) is a multivariate statistical technique that can be used to define subsystems of functionally connected brain regions based on the analysis of functional magnetic resonance imaging (fMRI) data. Here we introduce three-way multidimensional scaling as a method for the analysis of a group of fMRI data, which yields both a generic interregional configuration in low-dimensional space and a measure of each individual's deviation from the generic configuration. The distance between two generic interregional configurations obtained by MDS of two groups of data can be minimized by generalized Procrustes analysis, and the probability under the null hypothesis (that the two groups are sampled from the same population) of any residual group difference in interregional configurations can be assessed by a permutation test. These methods are developed and applied to activated fMRI time series acquired from 19 patients with schizophrenia and 20 normal comparison subjects during the performance of a semantic categorization and subvocal rehearsal task. The first three scaling dimensions are interpretable in terms of the major anatomical or functional subsystems of the activated system: "left-right," "input processing-other," and "subvocal output-other". We found no significant global or local differences between groups in interregional configurations in this 3D space. However, there was significantly greater variability of interregional configurations within the group of patients with schizophrenia. The implications for schizophrenia as a disconnexion disorder are discussed.