Dependence of cognitive ability on synchronous neural interactions determined by magnetoencephalography.

Previous studies have shown that synchronous neural interactions (SNIs) underlying healthy brain function can be readily distinguished from neural anomalies associated with diseases including dementia; however, it is imperative to identify biomarkers that facilitate early identification of individuals at risk for cognitive decline before the onset of clinical symptoms. Here, we evaluated whether variation in brain function, controlling for age, corresponds with subtle decrements in cognitive performance in cognitively healthy women. A total of 251 women (age range 24-102 yr) who performed above established cutoffs on the Montreal cognitive assessment (MoCA) also underwent a task-free magnetoencephalography scan from which SNIs were computed. The results demonstrated that increased SNI was significantly associated with decreased cognitive performance (r2 = 0.923, P = 0.009), controlling for age. Compared with the lowest performers with normal cognition (MoCA = 26), SNI of the highest performers (MoCA = 30) was associated with decorrelation primarily in the right anterior temporal cortex region, with additional (weaker) foci in left anterior temporal cortex, right posterior temporal cortex, and cerebellum. The findings highlight the relevance of neural network decorrelation on cognitive functioning and suggest that subtle increases in SNI may presage future cognitive impairment.NEW & NOTEWORTHY This study in cognitively healthy women showed that decreased cognitive performance is associated with increased neural network correlations, particularly involving the temporal cortices. As healthy brain function relies on dynamic neural network communication, these findings suggest that subtle increases in correlated neural network activity may be a useful early indicator of decrements in cognitive function.

Human Leukocyte Antigen (HLA) Modulates the Dependence on Age of the Variability of Synchronous Neural Interactions.

Recent evidence documented a protective effect of Class II human leukocyte antigen (HLA) DRB1*13 on brain health across the lifespan including evidence of reduced neural network variability relative to non-carriers. Here, in an extension of those findings, we evaluated the influence of a large number of Class I and Class II HLA alleles on aging-related changes in neural network variability. Cognitively healthy women (N = 178) ranging in age from 28 to 99 years old underwent a magnetoencephalography scan from which neural network variability was calculated and provided a blood sample from which HLA and apolipoprotein E (ApoE) genotype were determined. The primary analyses assessed the dependence of network variability on age in carriers of a specific HLA allele compared to non-carriers. Effects were considered protective if there was a significant increase of network variability with age in the absence of a given HLA allele but not in its presence, and were considered to confer susceptibility if the converse was documented; HLA alleles that did not influence the dependence of network variability on age in their presence or absence were considered neutral. Of 50 alleles investigated, 22 were found to be protective, 7 were found to confer susceptibility, and 21 were neutral. The frequencies of those 50 alleles were not associated significantly with ApoE genotype. The findings, which document the influence of HLA on age-related brain changes and highlight the role of HLA in healthy brain function, are discussed in terms of the role of HLA in the human immune response to foreign antigens.

MEG neural signature of sexual trauma in women veterans with PTSD.

Previous research has documented the utility of synchronous neural interactions (SNI) in classifying women veterans with and without posttraumatic stress disorder (PTSD) and other trauma-related outcomes based on functional connectivity using magnetoencephalography (MEG). Here, we extend that line of research to evaluate trauma-specific PTSD neural signatures with MEG in women veterans. Participants completed diagnostic interviews and underwent a task-free MEG scan from which SNI was computed. Thirty-five women veterans were diagnosed with PTSD due to sexual trauma and sixteen with PTSD due to non-sexual trauma. Strength of SNI was compared in women with and without sexual trauma, and linear discriminant analysis was used to classify the brain patterns of women with PTSD due to sexual trauma and non-sexual trauma. Comparison of SNI strength between the two groups revealed widespread hypercorrelation in women with sexual trauma relative to those without sexual trauma. Furthermore, using SNI, the brains of participants were classified as sexual trauma or non-sexual trauma with 100% accuracy. These findings bolster evidence supporting the utility of task-free SNI and suggest that neural signatures of PTSD are trauma-specific.

Classification of posttraumatic stress disorder and related outcomes in women veterans using magnetoencephalography.

Women veterans represent a unique population whose experiences and neurobiology differ from that of their male counterparts. Thus, while previous research has demonstrated the utility of synchronous neural interactions (SNI) as a biomarker of posttraumatic stress disorder (PTSD) in male veterans, the utility of SNI as a biomarker of PTSD in women veterans is unclear. Here we extend that line of research to evaluate classification of women veterans with and without PTSD and other trauma-related outcomes based on functional connectivity using magnetoencephalography (MEG). A total of 121 U.S. women veterans completed diagnostic interviews and underwent a task-free MEG scan from which SNI was computed. Linear discriminant analysis was used to classify PTSD and control groups according to SNI. That discriminant function was then used to classify each individual in the partial recovery and full recovery diagnostic groups as PTSD or control. All individuals were classified correctly (100% accuracy) according to their SNI in their PTSD and control groups. Seventy-seven percent of the full recovery group and 69% of the partial recovery group were classified as control. Individual staging in PTSD recovery was captured by the Mahalanobis D2 distances from the center of the control and PTSD centroid clusters. These findings provide compelling evidence supporting the utility of task-free SNI as a biomarker of PTSD and related outcomes in women veterans.

The effects of human leukocyte antigen DRB1*13 and apolipoprotein E on age-related variability of synchronous neural interactions in healthy women.

BACKGROUND: Age-related brain changes are well-documented and influenced by genetics. Extensive research links apolipoprotein E (apoE) to brain function, with the E4 allele serving as a risk factor for brain disease, including Alzheimer's disease, and the E2 allele conferring protection. Recent evidence also supports protective effects of another gene, human leukocyte antigen (HLA) DRB1*13, on brain disease and age-related brain atrophy in cognitively healthy adults. Here we investigated the effects of apoE and HLA DRB1*13 on brain function by examining changes in neural network properties with age in healthy adults. METHODS: One hundred seventy-eight cognitively healthy women (28-99 y old) underwent a magnetoencephalography scan and provided a blood sample for genetic analysis. Age-related changes in neural network variability in genetic subgroups of DRB1*13 × apoE genotype combinations were assessed using linear regression of network variability against age. FINDINGS: For individuals lacking a DRB1*13 allele and/or carrying an apoE4 allele, network variability increased significantly with age. In contrast, no such increase was observed in the presence of DRB1*13 and/or apoE2. INTERPRETATION: These findings extend previous research documenting the protective effect of DRB1*13 on brain structure to include protection against age-related changes in brain function, and demonstrate similar protective effects on neural network variability for either DRB1*13 or apoE2. These protective effects could be due to reduction or elimination of factors known to disrupt brain function, including neuroinflammation and amyloid beta protein. FUNDING: U.S. Department of Veterans Affairs, and University of Minnesota.

Brain Function in Gulf War Illness (GWI) and Associated Mental Health Comorbidities.

GWI has affected a substantial number of Gulf War (GW) veterans. The disease involves several organ systems among which the brain is most prominent. Neurological, cognitive and mood-related (NCM) symptoms frequently dominate and are at the root of chronic ill-health and disability in veterans suffering from GWI. In addition, such symptoms frequently co-occur with diagnosable mental health disorders, predominantly posttraumatic stress disorder (PTSD). Here we investigated the possibility that increased GWI severity leads, above a threshold, to a diagnosable mental health disorder (excluding psychosis). For this purpose, we used, in separate analyses, symptom severity scores and resting-state brain functional connectivity patterns, as determined by magnetoencephalography (MEG). Two-hundred-thirty GW-era veterans participated in this study. They completed diagnostic interviews to establish the presence of GWI and assess mental health status. This distinguished 3 groups: healthy controls (N = 41), veterans with GWI and no mental illness (GWI group, N = 91), and veterans with both GWI and mental health disorder (GWI+MH, N = 98). For each veteran, symptom severity scores in the 6 GWI domains (fatigue, pain, NCM, skin, gastrointestinal, respiratory) were available as well as 9 summary measures of the distribution of Synchronous Neural Interactions (SNI) derived from the MEG recordings. We tested the hypothesis that, in the presence of GWI, the appearance of a diagnosable mental health disorder may depend on GWI symptom severity. For that purpose, we performed a logistic regression on the GWI population, where the presence (or absence) of the MH disorder was the dependent variable and the age- and gender-adjusted GWI severity in the 6-symptom domains were the predictors. The outcome was the probability that a participant will have MH disorder or not. Similarly, we tested the hypothesis that the presence of the MH disorder can be predicted by the SNI distribution patterns by performing a second logistic regression as above but with the 9 SNI measures as predictors. We found GWI symptom severity differed significantly across groups (GWI+MH > GWI > Control). SNI distributions of the GWI group also differed significantly from the other groups in a systematic hemispheric pattern, such that the presence of GWI involved predominantly the left hemisphere, and presence of mental health disorders involved, in addition, the right hemisphere. Both logistic regressions yielded highly significant outcomes, demonstrating that both GWI symptom severity and SNI distribution measures can predict the presence of MH disorder in GWI. Remarkably, the prediction probabilities for MH presence derived from the symptom-based and SNI-based logistic regressions were positively and highly statistically significantly correlated. Taken together, both objective (neural) and subjective (symptoms) indices suggest that GWI is distinct from healthy controls and varies in severity in a continuum that leads, at the higher end, to a diagnosable MH disorder. The positive correlation between the GWI symptom-based and brain-based predicted classifications provides a key link between GWI symptom severity and synchronous neural interactions in the context of mental illness.

Gulf War illness (GWI) as a neuroimmune disease.

Gulf War illness (GWI) is a chronic disease characterized by the involvement of several organs, including the brain (Christova et al., Exp Brain Res doi: 10.1007/s00221-017-5010-8 , 2017). In a previous study (Georgopoulos et al., J Neural Eng 4:349-355, 2015), we identified six protective alleles from Class II human leukocyte antigen (HLA) genes, and more recently, we investigated the brain correlates of this protection (James et al., EBioMedicine 13:72-79, 2016). Those and other studies (Israeli, Lupus, 21:190-194, 2012) suggested an involvement of the immune system in GWI. In a recent study (Engdahl et al., EBioMedicine doi: 10.1016/j.ebiom.2016.08.030 , 2016), we showed that the brain pattern of synchronous neural interactions (SNI; Georgopoulos et al., J Neural Eng 4:349-355, 2007) in GWI is distinctly different from that in healthy controls. Here we focused on the SNI itself, as a basic measure of neural communication (irrespective of specific connections) and compared it between GWI and seven other diseases that cover a broad spectrum of etiology and pathophysiology. Specifically, we sought to determine which, if any, of those diseases might resemble GWI SNI, overall and within the HLA protective domain, and thus gain further knowledge regarding the nature of GWI brain abnormality. We studied a total of 962 participants from a healthy control population (N = 583) and eight different diseases, including GWI (N = 40), schizophrenia (SZ; N = 21), Alzheimer's disease (AD; N = 66), posttraumatic stress disorder (PTSD; N = 159), major depressive disorder (MDD; N = 10), relapsing-remitting multiple sclerosis (RRMS; N = 43), Sjögren's syndrome (SS; N = 32), and rheumatoid arthritis (RA; N = 8). They all underwent a resting-state magnetoencephalographic (MEG) scan to calculate SNIs. Data were analyzed using analysis of covariance (ANCOVA) with disease as fixed factor, and sex and age as covariates. We found that GWI SNIs differed significantly from control SZ, AD, PTSD and MDD but not from RRMS, SS and RA. In addition, we compared GWI to RRMS, SS and RA with respect to SNIs of MEG sensor pairs that were related to the HLA alleles protective for GWI (James et al., EBioMedicine 13:72-79, 2016). We found that GWI SNIs did not differ significantly from any of these three diseases but they did so from control SZ, AD, PTSD and MDD. These findings indicate that (a) GWI brain synchronicity does not differ significantly from that of known immune-related diseases (RRMS, SS, RA), and (b) that this SNI similarity is present within the HLA-related SNIs. In contrast, GWI SNIs differed significantly from those of the other diseases. We conclude that altered brain communication in GWI likely reflects immune-related processes, as postulated previously (James et al., EBioMedicine 13:72-79, 2016). By extension, these findings also indicate that functional brain abnormalities in RRMS, SS and RA might be, in part, due to lack of protective HLA alleles as documented for GWI (Georgopoulos et al., EBioMedicine 3:79-85, 2015).

Brain Correlates of Human Leukocyte Antigen (HLA) Protection in Gulf War Illness (GWI).

BACKGROUND: We recently reported that six alleles from class II genes of the Human Leukocyte Antigen (HLA) confer protection from Gulf War Illness (GWI) (Georgopoulos et al., 2015). The most significant effect is exerted on Neurological-Cognitive-Mood (NCM), Pain, and Fatigue symptoms, such that higher number of copies of the protective alleles are associated with lower symptom severity. Here we tested the hypothesis that this effect is exerted by modulating the strength of neural synchronicity. METHODS: Eighty-one Gulf War veterans (65 with GWI and 16 healthy controls) underwent a magnetoencephalography (MEG) scan to assess the strength of brain synchronicity by computing zero-lag crosscorrelations (and their Fisher z transforms) between prewhitened MEG time series. A high-resolution HLA genotyping determined the number of copies, k, of the 6 protective alleles above in each participant. We tested the hypothesis above by regressing NCM, Pain and Fatigue symptom severity against the interaction term, k×z (HLA-related effect), while including z (non-HLA-related effect), gender and age as covariates. The k×z and z terms assessed HLA- and non-HLA-related effects, respectively, of neural synchronicity on symptom severity. The distributions of these effects in sensor space were visualized using statistical heatmaps. FINDINGS: We found significant, graded HLA- and non-HLA-related effects: (a) NCM>Pain>Fatigue for HLA-related effects, (b) NCM>Fatigue>Pain for non-HLA-related effects, and (c) HLA-related>non-HLA-related effects for all symptoms. These effects had widespread but distinct distributions in sensor space that allowed the orderly separation of the 6 terms (3 symptom domains×2 HLA factors) in a multidimensional plot, where one dimension separated the symptoms and the other the HLA relation. INTERPRETATION: These findings demonstrate the presence of substantial, widespread, distinct and orderly HLA- and non-HLA-related neural influences on NCM, Pain and Fatigue symptom severity in GWI. FUNDING: U.S. Department of Veterans Affairs and University of Minnesota.

A Magnetoencephalographic (MEG) Study of Gulf War Illness (GWI).

BACKGROUND: Gulf War Illness (GWI) has affected many Gulf War veterans. It involves several organs, most notably the brain. Neurological-cognitive-mood-related symptoms frequently dominate and are at the root of chronic ill-health and disability in GWI. Here we investigated the neural mechanisms underlying brain dysfunction in GWI in the absence of mental health disorders. METHODS: Eighty-six veterans completed diagnostic interviews to establish the presence of GWI and assess mental health status. Participants diagnosed with GWI met both Center for Disease Control and Kansas criteria. We studied 46 healthy controls and 40 veterans with GWI without mental illness. They all underwent a resting-state magnetoencephalographic (MEG) scan to assess brain communication based on synchronous neural interactions (SNI; Georgopoulos et al., 2007). FINDINGS: We found substantial differences in SNI between control and GWI groups centered on the cerebellum and frontal cortex. In addition, using the maxima and minima of SNI per sensor as predictors, we successfully classified 94.2% of the 86 participants (95% sensitivity, 93.5% specificity). INTERPRETATION: These findings document distinct differences in brain function between control and GWI in the absence of mental health comorbidities, differences that are excellent predictors of GWI. FUNDING: U.S. Department of Veterans Affairs and University of Minnesota.

Cortical miscommunication after prenatal exposure to alcohol.

We report on the effects of prenatal alcohol exposure on resting-state brain activity as measured by magnetoencephalography (MEG). We studied 37 subjects diagnosed with fetal alcohol spectrum disorder in one of three categories: fetal alcohol syndrome, partial fetal alcohol syndrome, and alcohol-related neurodevelopmental disorder. For each subject, the MEG signal was recorded for 60 s during rest while subjects lay supine. Using time series analysis, we calculated the synchronous neural interactions for all pair-wise combinations of 248 MEG sensors resulting in 30,628 partial correlations for each subject. We found significant differences from control subjects in 6.19 % of the partial zero-lag crosscorrelations (synchronous neural interactions; Georgopoulos et al. in J Neural Eng 4:349-355, 2007), with these differences localized in the right posterior frontal, right parietal, and left parietal/posterior frontal regions. These results show that MEG can detect functional brain differences in the individuals affected by prenatal exposure to alcohol. Furthermore, these differences may serve as a biomarker for future studies linking symptoms and signs to specific brain areas. This may lead to new insights into the neuropathology of fetal alcohol spectrum disorders.

Differential neural activity patterns for spatial relations in humans: a MEG study.

Children learn the words for above-below relations earlier than for left-right relations, despite treating these equally well in a simple visual categorization task. Even as adults--conflicts in congruency, such as when a stimulus is depicted in a spatially incongruent manner with respect to salient global cues--can be challenging. Here we investigated the neural correlates of encoding and maintaining in working memory above-below and left-right relational planes in 12 adults using magnetoencephalography in order to discover whether above-below relations are represented by the brain differently than left-right relations. Adults performed perfectly on the task behaviorally, so any differences in neural activity were attributed to the stimuli's cognitive attributes. In comparing above-below to left-right relations during stimulus encoding, we found the greatest differences in neural activity in areas associated with space and movement. In comparing congruent to incongruent trials, we found the greatest differential activity in premotor areas. For both contrasts, brain areas involved in the encoding phase were also involved in the maintenance phase, which provides evidence that those brain areas are particularly important in representing the relational planes or congruency types throughout the trial. When comparing neural activity associated with the relational planes during working memory, additional right posterior areas were implicated, whereas the congruent-incongruent contrast implicated additional bilateral frontal and temporal areas. These findings are consistent with the hypothesis left-right relations are represented differently than above-below relations.

Pathological personality traits modulate neural interactions.

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), includes an empirically supported dimensional model of personality pathology that is assessed via the Personality Inventory for the DSM-5 (PID-5). Here we used magnetoencephalography (MEG; 248 sensors) to evaluate resting-state neural network properties associated with the five primary DSM-5 maladaptive personality domains (negative affect, detachment, antagonism, disinhibition, and psychoticism) in 150 healthy veterans ("control" group) and 179 veterans with various psychiatric disorders ("psychopathology" group). Since a fundamental network property is the strength of functional connectivity among network elements, we used the absolute value of the pairwise correlation coefficient (aCC) between prewhitened MEG sensor time series as a measure of neural functional connectivity and assessed its relations to the quantitative PID-5 scores in a linear regression model, where the log-transformed aCC was the dependent variable and individual PID scores, age, and gender were the independent variables. The partial regression coefficient (pRC) for a specific PID-5 score in that model provided information concerning the direction (positive, negative) and size (absolute value) of the PID effect on the strength of neural correlations. We found that, overall, PID domains had a negative effect (i.e., negative pRC; decorrelation) on aCC in the control group, but a positive one (i.e., positive pRC; hyper-correlation) in the psychopathology group. This dissociation of PID effects on aCC was especially pronounced for disinhibition, psychoticism, and negative affect. These results document for the first time a fundamental difference in neural-PID relations between control and psychopathology groups.

Neural mechanisms underlying the exploration of small city maps using magnetoencephalography.

The neural mechanisms underlying spatial cognition in the context of exploring realistic city maps are unknown. We conducted a novel brain imaging study to address the question of whether and how features of special importance for map exploration are encoded in the brain to make a spatial decision. Subjects explored by eyes small city maps exemplifying five different street network types in order to locate a hypothetical City Hall, while neural activity was recorded continuously by 248 magnetoencephalography (MEG) sensors at high temporal resolution. Monitoring subjects' eye positions, we locally characterized the maps by computing three spatial parameters of the areas that were explored. We computed the number of street intersections, the total street length, and the regularity index in the circular areas of 6 degrees of visual angle radius centered on instantaneous eye positions. We tested the hypothesis that neural activity during exploration is associated with the spatial parameters and modulated by street network type. All time series were rendered stationary and nonautocorrelated by applying an autoregressive integrated moving average model and taking the residuals. We then assessed the associations between the prewhitened time-varying MEG time series from 248 sensors and the prewhitened spatial parameters time series, for each street network type, using multiple linear regression analyses. In accord with our hypothesis, ongoing neural activity was strongly associated with the spatial parameters through localized and distributed networks, and neural processing of these parameters depended on the type of street network. Overall, processing of the spatial parameters seems to predominantly involve right frontal and prefrontal areas, but not for all street network layouts. These results are in line with findings from a series of previous studies showing that frontal and prefrontal areas are involved in the processing of spatial information and decision making. Modulation of neural processing of the spatial parameters by street network type suggests that some street network layouts may contain other types of spatial information that subjects use to explore maps and make spatial decisions.

Neural communication in posttraumatic growth.

Posttraumatic growth (PTG), or positive psychological changes following exposure to traumatic events, is commonly reported among trauma survivors. In the present study, we examined neural correlates of PTG in 106 veterans with PTSD and 193 veteran controls using task-free magnetoencephalography (MEG), diagnostic interviews and measures of PTG, and traumatic event exposure. Global synchronous neural interactions (SNIs) were significantly modulated downward with increasing PTG scores in controls (p = .005), but not in veterans with PTSD (p = .601). This effect was primarily characterized by negative slopes in local neural networks, was strongest in the medial prefrontal cortex, and was much stronger and more extensive in the control than the PTSD group. The present study complements previous research highlighting the role of neural adaptation in healthy functioning.

Development and application of a diagnostic algorithm for posttraumatic stress disorder.

Intact cognitive functions rely on synchronous neural activity; conversely, alterations in synchrony are thought to underlie psychopathology. We recently demonstrated that anomalies in synchronous neural interactions (SNI) determined by magnetoencephalography represent a putative PTSD biomarker. Here we develop and apply a regression-based diagnostic algorithm to further validate SNI as a PTSD biomarker in 432 veterans (235 controls; 138 pure PTSD; 59 PTSD plus comorbid disorders). Correlation coefficients served as proximities in multidimensional scaling (MDS) to obtain a two-dimensional representation of the data. In addition, least absolute shrinkage and selection operator (LASSO) regression was used to derive a diagnostic algorithm for PTSD. Performance of this algorithm was assessed by the area under the receiver operating characteristic (ROC) curves, sensitivity, and specificity in 1000 randomly divided testing and validation datasets and in independent samples. MDS revealed that individuals with PTSD, regardless of comorbid psychiatric conditions, are highly distinct from controls. Similarly, application of the LASSO regression-derived prediction model demonstrated remarkable classification accuracy (AUCs≥0.93 for men, AUC=0.82 for women). Neural functioning in individuals with PTSD, regardless of comorbid psychiatric diagnoses, can be used as a diagnostic test to determine patient disease status, further validating SNI as a PTSD biomarker.

Departure from Network Equilibrium (DNE): an efficient and scalable measure of instantaneous network dynamics, with an application to magnetoencephalography.

The assessment of the dynamic status of a network is currently unavailable. It is important to know how far a network is away from its equilibrium (as an indicator of instability) at a moment, and over periods of time. Here, we introduce the Departure from Network Equilibrium (DNE), a new measure of instantaneous network dynamics. DNE is simple, fast to compute, and scalable with network size. We present the results of its application on white noise networks (as a basis) and on networks derived from magnetoencephalographic recordings from the human brain.

The number of cysteine residues per mole in apolipoprotein E affects systematically synchronous neural interactions in women's healthy brains.

Apolipoprotein E (apoE) is involved in lipid metabolism in the brain, but its effects on brain function are not understood. Three apoE isoforms (E4, E3, and E2) are the result of cysteine-arginine interchanges at two sites: there are zero interchanges in E4, one interchange in E3, and two interchanges in E2. The resulting six apoE genotypes (E4/4, E4/3, E4/2, E3/3, E3/2, E2/2) yield five groups with respect to the number of cysteine residues per mole (CysR/mole), as follows. ApoE4/4 has zero cysteine residues per mole (0-CysR/mole), E4/3 has one (1-CysR/mole), E4/2 and E3/3 each has two (2-CysR/mole), E3/2 has three (3-CysR/mole), and E2/2 has four (4-CysR/mole). The use of the number of CysR/mole to characterize the apoE molecule converts the categorical apoE genotype scale, consisting of 6 distinct genotypes above, to a 5-point continuous scale (0-4 CysR/mole). This allows the use of statistical analyses suitable for continuous variables (e.g. regression) to quantify the relations between various variables and apoE. Using such analyses, here, we show for the first time that apoE affects in a graded and orderly manner neural communication, as assessed by analyzing the relation between the number of CysR/mole and synchronous neural interactions (SNI) measured by magnetoencephalography (MEG) in 130 cognitively healthy women. At the one end of the CysR/mole range, the 4-CysR/mole (E2/2) SNI distribution had the highest mean, lowest variance, lowest range, and lowest coefficient of variation, whereas at the other end, 0-CysR/mole (E4/4) SNI distribution had the lowest mean, highest variance, highest range, and highest coefficient of variation. The special status of the 4-CysR/mole distribution was reinforced by the results of a hierarchical tree analysis where the 4-CysR/mole (E2/2) SNI distribution occupied a separate branch by itself and the remaining CysR/mole SNI distributions were placed at increasing distances from the 4-CysR/mole distribution, according to their number of CysR/mole, with the 0-CysR/mole (E4/4) being farthest away. These findings suggest that the 4-CysR/mole (E2/2) SNI distribution could serve as a reference distribution. When the SNI distributions of individual women were expressed as distances from this reference distribution, there was a substantial overlap among women of various CysR/mole. This refocuses the placement of individual brains along a continuous distance from the 4-CysR/mole SNI distribution, in contrast to the common categorical assignment to a specific apoE genotype. Finally, the orderly variation of SNI with the number of CysR/mole found here is in keeping with recent advances and ideas regarding the molecular mechanisms underlying the differential effects of apoE in the brain which emphasize the healthier stability conferred on the apoE molecule by the increasing number of cysteine-arginine interchanges, with 4-CysR/mole (E2/2) being the best case, as opposed to the instability and increased chance of toxic fragmentation of the apoE molecule with lower number of CysR/mole, with 0-CysR/mole (E4/4) as the worst case (Mahley and Huang in Neuron 76:871-885, 2012a). However, our results also document the appreciable variation of SNI properties within the various CysR/mole groups and individuals which points to the existence and important role of other factors involved in shaping brain function at the network level.

Canonical correlation analysis of synchronous neural interactions and cognitive deficits in Alzheimer's dementia.

In previous work (Georgopoulos et al 2007 J. Neural Eng. 4 349-55) we reported on the use of magnetoencephalographic (MEG) synchronous neural interactions (SNI) as a functional biomarker in Alzheimer's dementia (AD) diagnosis. Here we report on the application of canonical correlation analysis to investigate the relations between SNI and cognitive neuropsychological (NP) domains in AD patients. First, we performed individual correlations between each SNI and each NP, which provided an initial link between SNI and specific cognitive tests. Next, we performed factor analysis on each set, followed by a canonical correlation analysis between the derived SNI and NP factors. This last analysis optimally associated the entire MEG signal with cognitive function. The results revealed that SNI as a whole were mostly associated with memory and language, and, slightly less, executive function, processing speed and visuospatial abilities, thus differentiating functions subserved by the frontoparietal and the temporal cortices. These findings provide a direct interpretation of the information carried by the SNI and set the basis for identifying specific neural disease phenotypes according to cognitive deficits.

Spatiotemporal neural interactions underlying continuous drawing movements as revealed by magnetoencephalography.

Continuous and sequential movements are controlled by widely distributed brain regions. A series of studies have contributed to understanding the functional role of these regions in a variety of visuomotor tasks. However, little is known about the neural interactions underpinning continuous movements. In the current study, we examine the spatiotemporal neural interactions underlying continuous drawing movements and the association of them with behavioral components. We conducted an experiment in which subjects copied a pentagon continuously for ~45 s using an XY joystick, while neuromagnetic fluxes were recorded from their head using a 248-sensor whole-head magnetoencephalography (MEG) device. Each sensor time series was rendered stationary and non-autocorrelated by applying an autoregressive integrated moving average model and taking the residuals. We used the directional variability of the movement as a behavioral measure of the controls generated. The main objective of this study was to assess the relation between neural interactions and the variability of movement direction. That is, we divided the continuous recordings into consecutive periods (i.e., time-bins) of 51 steps duration and computed the pairwise cross-correlations between the prewhitened time series in each time-bin. The circular standard deviation of the movement direction within each time-bin provides an estimate of the directional variability of the 51-ms trajectory segment. We looked at the association between neural interactions and variability of movement direction, separately for each pair of sensors, by running a cross-correlation analysis between the strength of the MEG pairwise cross-correlations and the circular standard deviations. We identified two types of neuronal networks: in one, the neural interactions are correlated with the directional variability of the movement at negative time-lags (feedforward), and in the other, the neural interactions are correlated with the directional variability of the movement at positive time-lags (feedback). Sensors associated mostly with feedforward processes are distributed in the left hemisphere and the right occipital-temporal junction, whereas sensors related to feedback processes are distributed in the right hemisphere and the left cerebellar hemisphere. These results are in line with findings from a series of previous studies showing that specific brain regions are involved in feedforward and feedback control processes to plan, perform, and correct movements. Additionally, we looked at whether changes in movement direction modulate the neural interactions. Interestingly, we found a preponderance of sensors associated with changes in movement direction over the right hemisphere-ipsilateral to the moving hand. These sensors exhibit stronger coupling with the rest of the sensors for trajectory segments with high rather than low directional movement variability. We interpret these results as evidence that ipsilateral cortical regions are recruited for continuous movements when the curvature of the trajectory increases. To the best of our knowledge, this is the first study that shows how neural interactions are associated with a behavioral control parameter in continuous and sequential movements.

The temporal, spatial, and frequency dimensions of neural oscillations associated with verbal working memory.

Cognitive operations engage neural generators oscillating at different frequencies distributed in time and space. Accordingly, oscillatory activity detected by magnetoencephalography (MEG)/electroencephalography (EEG) should be analyzed along frequency, time, and spatial dimensions. MEG data were obtained from 19 healthy individuals while performing a modified Sternberg paradigm. The stimuli were letters, which constituted words or pronounceable nonwords. We applied tridimensional analysis of oscillations and also computed event-related fields (ERFs) in areas where significant changes in oscillatory activity were observed. Verbal working memory for visual verbal stimuli was associated with oscillatory interplay between the bilateral occipital lobes and the left frontoparietotemporal areas. Spatially stable occipital desynchrony was noted during information encoding, while a left hemisphere desynchronization, increasing in amplitude and spatial extent over time, was observed during information encoding and maintenance. No ERF changes were detected during information maintenance. Oscillatory activity associated with verbal working memory is consistent with the above hypothesis. These findings underscore the importance of multidimensional evaluation of oscillations. The findings also indicate that combining electrophysiological methods increase the chance of signal detection.