Narrowband multivariate source separation for semi-blind discovery of experiment contrasts.

BACKGROUND: Electrophysiological recordings contain mixtures of signals from distinct neural sources, impeding a straightforward interpretation of the sensor-level data. This mixing is particularly detrimental when distinct sources resonate in overlapping frequencies. Fortunately, the mixing is linear and instantaneous. Multivariate source separation methods may therefore successfully separate statistical sources, even with overlapping spatial distributions. NEW METHOD: We demonstrate a feature-guided multivariate source separation method that is tuned to narrowband frequency content as well as binary condition differences. This method - comparison scanning generalized eigendecomposition, csGED - harnesses the covariance structure of multichannel data to find directions (i.e., eigenvectors) that maximally separate two subsets of data. To drive condition specificity and frequency specificity, our data subsets were taken from different task conditions and narrowband-filtered prior to applying GED. RESULTS: To validate the method, we simulated MEG data in two conditions with shared noise characteristics and unique signal. csGED outperformed the best sensor at reconstructing the ground truth signals, even in the presence of large amounts of noise. We next applied csGED to a published empirical MEG dataset on visual perception vs. imagery. csGED identified sources in alpha, beta, and gamma bands, and successfully separated distinct networks in the same frequency band. COMPARISON WITH EXISTING METHOD(S): GED is a flexible feature-guided decomposition method that has previously successfully been applied. Our combined frequency- and condition-tuning is a novel adaptation that extends the power of GED in cognitive electrophysiology. CONCLUSIONS: We demonstrate successful condition-specific source separation by applying csGED to simulated and empirical data.

Attention and temporal expectations modulate power, not phase, of ongoing alpha oscillations.

The perception of near-threshold visual stimuli has been shown to depend in part on the phase (i.e., time in the cycle) of ongoing alpha (8-13 Hz) oscillations in the visual cortex relative to the onset of that stimulus. However, it is currently unknown whether the phase of the ongoing alpha activity can be manipulated by top-down factors such as attention or expectancy. Using three variants of a cross-modal attention paradigm with constant predictable stimulus onsets, we examined if cues signaling to attend to either the visual or the auditory domain influenced the phase of alpha oscillations in the associated sensory cortices. Importantly, intermixed in all three experiments, we included trials without a target to estimate the phase at target presentation without contamination from the early evoked responses. For these blank trials, at the time of expected target and distractor onset, we examined (1) the degree of the uniformity in phase angles across trials, (2) differences in phase angle uniformity compared with a pretarget baseline, and (3) phase angle differences between visual and auditory target conditions. Across all three experiments, we found that, although the cues induced a modulation in alpha power in occipital electrodes, neither the visual condition nor the auditory cue condition induced any significant phase-locking across trials during expected target or distractor presentation. These results suggest that, although alpha power can be modulated by top-down factors such as attention and expectation, the phase of the ongoing alpha oscillation is not under such control.

Frontal oscillatory dynamics predict feedback learning and action adjustment.

Frontal oscillatory dynamics in the theta (4-8 Hz) and beta (20-30 Hz) frequency bands have been implicated in cognitive control processes. Here we investigated the changes in coordinated activity within and between frontal brain areas during feedback-based response learning. In a time estimation task, participants learned to press a button after specific, randomly selected time intervals (300-2000 msec) using the feedback after each button press (correct, too fast, too slow). Consistent with previous findings, theta-band activity over medial frontal scalp sites (presumably reflecting medial frontal cortex activity) was stronger after negative feedback, whereas beta-band activity was stronger after positive feedback. Theta-band power predicted learning only after negative feedback, and beta-band power predicted learning after positive and negative feedback. Furthermore, negative feedback increased theta-band intersite phase synchrony (a millisecond resolution measure of functional connectivity) among right lateral prefrontal, medial frontal, and sensorimotor sites. These results demonstrate the importance of frontal theta- and beta-band oscillations and intersite communication in the realization of reinforcement learning.

Error-related medial frontal theta activity predicts cingulate-related structural connectivity.

Studies on electrophysiological signatures of error processing have focused on the medial frontal cortex, although widespread neuroanatomical networks support error/action monitoring. Here, electrophysiological responses to errors were combined with structural white matter diffusion tensor imaging (DTI) to investigate the long-range anatomical networks that support error processing. The approach taken here was to link individual differences in error-related EEG responses to individual differences in white matter connectional anatomy. Twenty subjects performed a speeded instructed choice task (a variant of the Simon task) designed to elicit response errors, and also underwent DTI scanning in a separate session. In the EEG data, significantly enhanced theta (4-8 Hz) oscillations were observed over medial frontal electrodes (centered on FCz) during response errors. Mid-frontal scalp sites (likely reflecting medial frontal cortex activity) also functioned as a strong "hub" for information flow, measured through theta-band phase synchronization degree. Next, a dipole source of the error-related theta-band activity was localized for each subject, accounting for approximately 80% of the topographical variance. Correlating individual differences in medial frontal theta dynamics with white matter tracts linking these dipole sources to the rest of the brain revealed that subjects with stronger error-related theta also had stronger white matter connectivity with the ventral striatum and inferior frontal gyrus. Further, subjects in whom medial frontal regions acted as a stronger synchronization "hub" had stronger connectivity between the dipole source location and the corpus callosum and dorsomedial prefrontal white matter pathways. These findings provide novel evidence for the role of widespread fronto-striatal networks in monitoring actions and signaling behavioral errors.

Activation of the caudal anterior cingulate cortex due to task-related interference in an auditory Stroop paradigm.

Successful information processing requires the focusing of attention on a certain stimulus property and the simultaneous suppression of irrelevant information. The Stroop task is a useful paradigm to study such attentional top-down control in the presence of interference. Here, we investigated the neural correlates of an auditory Stroop task using fMRI. Subjects focused either on tone pitch (relatively high or low; phonetic task) or on the meaning of a spoken word (high/low/good; semantic task), while ignoring the other stimulus feature. We differentiated between task-related (phonetic incongruent vs. semantic incongruent) and sensory-level interference (phonetic incongruent vs. phonetic congruent). Task-related interference activated similar regions as in visual Stroop tasks, including the anterior cingulate cortex (ACC) and the presupplementary motor-area (pre-SMA). More specifically, we observed that the very caudal/posterior part of the ACC was activated and not the dorsal/anterior region. Because identical stimuli but different task demands are compared in this contrast, it reflects conflict at a relatively high processing level. A more conventional contrast between incongruent and congruent phonetic trials was associated with a different cluster in the pre-SMA/ACC which was observed in a large number of previous studies. Finally, functional connectivity analysis revealed that activity within the regions activated in the phonetic incongruent vs. semantic incongruent contrast was more strongly interrelated during semantically vs. phonetically incongruent trials. Taken together, we found (besides activation of regions well-known from visual Stroop tasks) activation of the very caudal and posterior part of the ACC due to task-related interference in an auditory Stroop task.

Different neural systems adjust motor behavior in response to reward and punishment.

Individuals use the outcomes of their actions to adjust future behavior. However, it remains unclear whether the same neural circuits are used to adjust behavior due to rewarding and punishing outcomes. Here we used functional magnetic resonance imaging (fMRI) and a reward-providing reaction time task to investigate the adaptation of a simple motor response following four different outcomes (delivery versus omission and monetary gain versus loss). We found that activation in the thalamus and insula predicted adjustments of motor responses due to outcomes that were cued and delivered, whereas activation in the ventral striatum predicted such adjustments when outcomes were cued but omitted. Further, activation of OFC predicted improvement after all punishing outcomes, independent of whether they were omitted rewards or delivered punishments. Finally, we found that activity in anterior cingulate predicted adjustment after delivered punishments and activity in dorsal striatum predicted adaptation after delivered rewards. Our results provide evidence that different but somewhat overlapping circuits mediate the same behavioral adaptation when it is driven by different incentive outcomes.

Reward expectation modulates feedback-related negativity and EEG spectra.

The ability to evaluate outcomes of previous decisions is critical to adaptive decision-making. The feedback-related negativity (FRN) is an event-related potential (ERP) modulation that distinguishes losses from wins, but little is known about the effects of outcome probability on these ERP responses. Further, little is known about the frequency characteristics of feedback processing, for example, event-related oscillations and phase synchronizations. Here, we report an EEG experiment designed to address these issues. Subjects engaged in a probabilistic reinforcement learning task in which we manipulated, across blocks, the probability of winning and losing to each of two possible decision options. Behaviorally, all subjects quickly adapted their decision-making to maximize rewards. ERP analyses revealed that the probability of reward modulated neural responses to wins, but not to losses. This was seen both across blocks as well as within blocks, as learning progressed. Frequency decomposition via complex wavelets revealed that EEG responses to losses, compared to wins, were associated with enhanced power and phase coherence in the theta frequency band. As in the ERP analyses, power and phase coherence values following wins but not losses were modulated by reward probability. Some findings between ERP and frequency analyses diverged, suggesting that these analytic approaches provide complementary insights into neural processing. These findings suggest that the neural mechanisms of feedback processing may differ between wins and losses.