Shared and specific independent components analysis for between-group comparison.

Independent component analysis (ICA) has been extensively used in individual and within-group data sets in real-world applications, but how can it be employed in a between-groups or conditions design? Here, we propose a new method to embed group membership information into the FastICA algorithm so as to extract components that are either shared between groups or specific to one or a subset of groups. The proposed algorithm is designed to automatically extract the pattern of differences between different experimental groups or conditions. A new constraint is added to the FastICA algorithm to simultaneously deal with the data of multiple groups in a single ICA run. This cost function restricts the specific components of one group to be orthogonal to the subspace spanned by the data of the other groups. As a result of performing a single ICA on the aggregate data of several experimental groups, the entire variability of data sets is used to extract the shared components. The results of simulations show that the proposed algorithm performs better than the regular method in both the reconstruction of the source signals and classification of shared and specific components. Also, the sensitivity to detect variations in the amplitude of shared components across groups is enhanced. A rigorous proof of convergence is provided for the proposed iterative algorithm. Thus, this algorithm is guaranteed to extract and classify shared and specific independent components across different experimental groups and conditions in a systematic way.

Functional connectivity in patients with idiopathic generalized epilepsy.

PURPOSE: Idiopathic generalized epilepsy (IGE) is characterized by electroencephalography (EEG) recordings with generalized spike wave discharges (GSWDs) arising from normal background activity. Although GSWDs are the result of highly synchronized activity in the thalamocortical network, EEG without GSWDs is believed to represent normal brain activity. The aim of this study was to investigate whether thalamocortical interactions are altered even during GSWD-free EEG periods in patients with IGE. METHODS: A GSWD-related group analysis was performed in 12 IGE patients to define seeds in areas involved during GSWDs. EEG-functional magnetic resonance imaging (fMRI) datasets from 22 IGE patients without GSWDs during the investigation and 30 age-matched healthy controls were then selected to investigate functional connectivity in GSWD-related areas. Blood oxygen level dependent (BOLD) signal changes were extracted from seeds defined by the GSWD-related group analysis. The averaged time course within each seed was used to detect brain regions with BOLD signal correlated with the seed. Group differences between patients and controls were estimated. KEY FINDINGS: The GSWD-related group analysis showed BOLD activation in the thalamus, the frontomesial cortex, and the cerebellum and BOLD deactivation in default mode areas. For the connectivity analysis, eight seeds were placed bilaterally in the thalamus, mesial frontal cortex, precuneus, and cerebellum. The functional connectivity analysis of these seeds did not show clearly altered functional connectivity for patients versus controls. SIGNIFICANCE: The results underscore the paroxysmal nature of GSWDs: Although GSWDs are characterized by highly synchronized activity in the thalamocortical network, the functional connectivity in areas involved during GSWDs does not demonstrate abnormality in GSWD-free periods.

Ten kilohertz SCS for Treatment of Chronic Upper Extremity Pain (UEP): Results from Prospective Observational Study.

BACKGROUND: Chronic upper extremity pain (UEP) has complex etiologies and is often disabling. It has been shown that 10 kHz SCS can provide paresthesia-free and durable pain relief in multiple pain types and improve the quality of life of patients. OBJECTIVE: To gain additional evidence on the safety and effectiveness of 10 kHz SCS for the treatment of chronic UEP. STUDY DESIGN: It was a prospective, multicenter, and observational study. The study was registered on ClinicalTrials.gov prospectively (clinical trial identifier: NCT02703818). SETTING: Multicenter. PATIENTS INTERVENTION AND MAIN OUTCOMES: A total of 43 subjects with chronic UEP of ≥5 cm (on a 0-10 cm visual analog scale; VAS) underwent a trial of 10 kHz SCS, and subjects with ≥40% pain relief received a permanent implant. All subjects had upper limb pain at baseline, while some had concomitant shoulder or neck pain. Subject outcomes were assessed for 12 months, and the primary outcome was the responder rate (percentage of subjects experiencing ≥50% pain relief from baseline) at three months. RESULTS: Thirty-eight subjects successfully completed the trial (88.3% success rate), 33 received permanent implants (five withdrew consent), and 32 had device activation (per protocol population). There were no paresthesias or uncomfortable changes in stimulation related to changes in posture during the study and there were no neurological deficits. Responder rates at 12 months for upper limb, shoulder, and neck pain in per protocol population (N=32) were 78.1%, 85.2%, and 75.0%, respectively. At 12 months, 84.4% of subjects were satisfied or very satisfied with 10 kHz SCS, and 38.7% either reduced or eliminated opioid usage. CONCLUSION: This study further supports the effectiveness of 10 kHz SCS for chronic UEP treatment and documents the safety profile of the therapy. CLINICAL TRIAL IDENTIFIER: NCT02703818.

Validation of Shared and Specific Independent Component Analysis (SSICA) for Between-Group Comparisons in fMRI.

Independent component analysis (ICA) has been widely used to study functional magnetic resonance imaging (fMRI) connectivity. However, the application of ICA in multi-group designs is not straightforward. We have recently developed a new method named "shared and specific independent component analysis" (SSICA) to perform between-group comparisons in the ICA framework. SSICA is sensitive to extract those components which represent a significant difference in functional connectivity between groups or conditions, i.e., components that could be considered "specific" for a group or condition. Here, we investigated the performance of SSICA on realistic simulations, and task fMRI data and compared the results with one of the state-of-the-art group ICA approaches to infer between-group differences. We examined SSICA robustness with respect to the number of allowable extracted specific components and between-group orthogonality assumptions. Furthermore, we proposed a modified formulation of the back-reconstruction method to generate group-level t-statistics maps based on SSICA results. We also evaluated the consistency and specificity of the extracted specific components by SSICA. The results on realistic simulated and real fMRI data showed that SSICA outperforms the regular group ICA approach in terms of reconstruction and classification performance. We demonstrated that SSICA is a powerful data-driven approach to detect patterns of differences in functional connectivity across groups/conditions, particularly in model-free designs such as resting-state fMRI. Our findings in task fMRI show that SSICA confirms results of the general linear model (GLM) analysis and when combined with clustering analysis, it complements GLM findings by providing additional information regarding the reliability and specificity of networks.

Specific resting-state brain networks in mesial temporal lobe epilepsy.

We studied with functional magnetic resonance imaging (fMRI) differences in resting-state networks between patients with mesial temporal lobe epilepsy (MTLE) and healthy subjects. To avoid any a priori hypothesis, we use a data-driven analysis assessing differences between groups independently of structures involved. Shared and specific independent component analysis (SSICA) is an exploratory method based on independent component analysis, which performs between-group network comparison. It extracts and classifies components (networks) in those common between groups and those specific to one group. Resting fMRI data were collected from 10 healthy subjects and 10 MTLE patients. SSICA was applied multiple times with altered initializations and different numbers of specific components. This resulted in many components specific to patients and to controls. Spatial clustering identified the reliable resting-state networks among all specific components in each group. For each reliable specific network, power spectrum analysis was performed on reconstructed time-series to estimate connectivity in each group and differences between groups. Two reliable networks, corresponding to statistically significant clusters robustly detected with clustering were labeled as specific to MTLE and one as specific to the control group. The most reliable MTLE network included hippocampus and amygdala bilaterally. The other MTLE network included the postcentral gyri and temporal poles. The control-specific network included bilateral precuneus, anterior cingulate, thalamus, and parahippocampal gyrus. Results indicated that the two MTLE networks show increased connectivity in patients, whereas the control-specific network shows decreased connectivity in patients. Our findings complement results from seed-based connectivity analysis (1). The pattern of changes in connectivity between mesial temporal lobe structures and other areas may help us understand the cognitive impairments often reported in patients with MTLE.

Resting-state connectivity of the sustained attention network correlates with disease duration in idiopathic generalized epilepsy.

INTRODUCTION: In idiopathic generalized epilepsy (IGE), a normal electroencephalogram between generalized spike and wave (GSW) discharges is believed to reflect normal brain function. However, some studies indicate that even excluding GSW-related errors, IGE patients perform poorly on sustained attention task, the deficit being worse as a function of disease duration. We hypothesized that at least in a subset of structures which are normally involved in sustained attention, resting-state functional connectivity (FC) is different in IGE patients compared to controls and that some of the changes are related to disease duration. METHOD: Seeds were selected based on a sustained attention study in controls. Resting-state functional magnetic resonance imaging (fMRI) data was obtained from 14 IGE patients and 14 matched controls. After physiological noise removal, the mean time-series of each seed was used as a regressor in a general linear model to detect regions that showed correlation with the seed. In patients, duration factor was defined based on epilepsy duration. Between-group differences weighted by the duration factor were evaluated with mixed-effects model. Correlation was then evaluated in IGE patients between the FC, averaged over each significant cluster, and the duration factor. RESULTS: Eight of 18 seeds showed significant difference in FC across groups. However, only for seeds in the medial superior frontal and precentral gyri and in the medial prefrontal area, average FC taken over significant clusters showed high correlation with the duration factor. These 3 seeds showed changes in FC respectively with the premotor and superior frontal gyrus, the dorsal premotor, and the supplementary motor area plus precentral gyrus. CONCLUSION: Alterations of FC in IGE patients are not limited to the frontal areas. However, as indicated by specificity analysis, patients with long history of disease show changes in FC mainly within the frontal areas.