Catherine Poulsen

Electrophysiological Responses to Errors and Feedback in the Process of Action Regulation

The anterior cingulate cortex (ACC) is believed to be involved in the executive control of actions, such as in monitoring conflicting response demands, detecting errors, and evaluating the emotional significance of events. In this study, participants performed a task in which evaluative feedback was delayed, so that it was irrelevant to immediate response control but retained its emotional value as a performance indicator. We found that a medial frontal feedback-related negativity similar to the error-related negativity (ERN) tracked affective response to the feedback and predicted subsequent performance. Source analysis of the feedback-related negativity and ERN revealed a common dorsomedial ACC source and a rostromedial ACC source specific to the ERN. The oscillatory nature of these sources provides further evidence that the ERN reflects ongoing theta activity generated in the mediofrontal regions. These results suggest that action regulation by the cingulate gyrus may require the entrainment of multiple structures of the Papez corticolimbic circuit.

Frontolimbic response to negative feedback in clinical depression

Functional neuroimaging suggests that limbic regions of the medial frontal cortex may be abnormally active in individuals with depression. These regions, including the anterior cingulate cortex, are engaged in both action regulation, such as monitoring errors and conflict, and affect regulation, such as responding to pain. The authors examined whether clinically depressed subjects would show abnormal sensitivity of frontolimbic networks as they evaluated negative feedback. Depressed subjects and matched control subjects performed a video game in the laboratory as a 256-channel EEG was recorded. Speed of performance on each trial was graded with a feedback signal of A, C, or F. By 350 ms after the feedback signal, depressed subjects showed a larger medial frontal negativity for all feedback compared with control subjects with a particularly striking response to the F grade. This response was strongest for moderately depressed subjects and was attenuated for subjects who were more severely depressed. Localization analyses suggested that negative feedback engaged sources in the anterior cingulate and insular cortices. These results suggest that moderate depression may sensitize limbic networks to respond strongly to aversive events.

EEG source localization: Sensor density and head surface coverage

BACKGROUND The accuracy of EEG source localization depends on a sufficient sampling of the surface potential field, an accurate conducting volume estimation (head model), and a suitable and well-understood inverse technique. The goal of the present study is to examine the effect of sampling density and coverage on the ability to accurately localize sources, using common linear inverse weight techniques, at different depths. Several inverse methods are examined, using the popular head conductivity. NEW METHOD Simulation studies were employed to examine the effect of spatial sampling of the potential field at the head surface, in terms of sensor density and coverage of the inferior and superior head regions. In addition, the effects of sensor density and coverage are investigated in the source localization of epileptiform EEG. RESULTS Greater sensor density improves source localization accuracy. Moreover, across all sampling density and inverse methods, adding samples on the inferior surface improves the accuracy of source estimates at all depths. COMPARISON WITH EXISTING METHODS More accurate source localization of EEG data can be achieved with high spatial sampling of the head surface electrodes. CONCLUSIONS The most accurate source localization is obtained when the voltage surface is densely sampled over both the superior and inferior surfaces.

Frontolimbic Activity and Cognitive Bias in Major Depression

In order to explore neural activity that accompanies cognitive bias in mood disorders, the authors had clinically depressed and nondepressed controls complete a self-evaluation procedure in which they indicated whether trait words were self-descriptive. Dense-array (256-channel) electroencephalography was recorded. Greater depression and low Positive Affect were associated with decreased endorsement of favorable (Good) traits, and greater anxiety and high Negative Affect were associated with increased endorsement of unfavorable (Bad) traits. For controls, the event-related potential (ERP) showed an enhanced visual N1 for trials in which Bad traits were endorsed. For depressed participants, this N1 was attenuated, specifically for these endorsed Bad trials. A similar pattern was observed in the P2-medial frontal negativity (P2-MFN) complex, with controls showing an enhanced MFN to the endorsed Bad words, while depressed participants showed an attenuated or absent medial frontal response on these items specifically. Distributed linear-inverse source analysis of the ERP localized the N1 effect to the inferotemporal-occipital cortex and the medial frontal effect to the dorsal anterior cingulate cortex. The altered ERP responses in depressed participants may provide clues to the neurophysiological processes associated with negatively biased cognition and self-evaluation in clinical depression.

Reference-free removal of EEG-fMRI ballistocardiogram artifacts with harmonic regression

Combining electroencephalogram (EEG) recording and functional magnetic resonance imaging (fMRI) offers the potential for imaging brain activity with high spatial and temporal resolution. This potential remains limited by the significant ballistocardiogram (BCG) artifacts induced in the EEG by cardiac pulsation-related head movement within the magnetic field. We model the BCG artifact using a harmonic basis, pose the artifact removal problem as a local harmonic regression analysis, and develop an efficient maximum likelihood algorithm to estimate and remove BCG artifacts. Our analysis paradigm accounts for time-frequency overlap between the BCG artifacts and neurophysiologic EEG signals, and tracks the spatiotemporal variations in both the artifact and the signal. We evaluate performance on: simulated oscillatory and evoked responses constructed with realistic artifacts; actual anesthesia-induced oscillatory recordings; and actual visual evoked potential recordings. In each case, the local harmonic regression analysis effectively removes the BCG artifacts, and recovers the neurophysiologic EEG signals. We further show that our algorithm outperforms commonly used reference-based and component analysis techniques, particularly in low SNR conditions, the presence of significant time-frequency overlap between the artifact and the signal, and/or large spatiotemporal variations in the BCG. Because our algorithm does not require reference signals and has low computational complexity, it offers a practical tool for removing BCG artifacts from EEG data recorded in combination with fMRI.

Critical periods for the neurodevelopmental processes of externalizing and internalizing

Research on neurobiological development is providing insight into the nature and mechanisms of human neural plasticity. These mechanisms appear to support two different forms of developmental learning. One form of learning could be described as externalizing, in which neural representations are highly responsive to environmental influences, as the child typically operates under a mode of hedonic approach. A second form of learning supports internalizing, in which motive control separates attention and self-regulation from the immediate influences of the context, particularly when the child faces conditions of avoidance and threat. The dorsal cortical networks of externalizing are organized around dorsal limbic (cingulate, septal, lateral hypothalamic, hippocampal, and ventral striatal) circuits. In contrast, the ventral cortical networks of internalizing are organized around ventral limbic (anterior temporal and orbital cortex, extended amygdala, dorsal striatal, and mediodorsal thalamic) circuits. These dual divisions of the limbic system in turn self-regulate their arousal levels through different brain stem and forebrain neuromodulator projection systems, with dorsal corticolimbic networks regulated strongly by locus coeruleus norepinephrine and brain stem raphe nucleus serotonin projection systems, and ventral corticolimbic networks regulated by ventral tegmental dopamine and forebrain acetylcholine projections. Because the arousal control systems appear to regulate specific properties of neural plasticity in development, an analysis of these systems explains differences between externalizing and internalizing at multiple levels of neural and psychological self-regulation. In neuroscience, the concept of critical periods has been applied to times when experience is essential for the maturation of sensory systems. In a more general neuropsychological analysis, certain periods of the childs development require successful self-regulation through the differential capacities for externalizing and internalizing.

Learning and the Development of Contexts for Action

Neurophysiological evidence from animal studies suggests that frontal corticolimbic systems support early stages of learning, whereas later stages involve context representation formed in hippocampus and posterior cingulate cortex. In dense-array EEG studies of human learning, we observed brain activity in medial prefrontal cortex (the medial frontal negativity or MFN) was not only observed in early stages, but, surprisingly, continued to increase as learning progressed. In the present study we investigated this finding by examining MFN amplitude as participants learned an arbitrary associative learning task over three sessions. On the fourth session the same task with new stimuli was presented to assess changes in MFN amplitude. The results showed that MFN amplitude continued to increase with practice over the first three sessions, in contrast to P3 amplitudes. Even when participants were presented with new stimuli in session 4, MFN amplitude was larger than that observed in the first session. Furthermore, MFN activity from the third session predicted learning rate in the fourth session. The results point to an interaction between early and late stages in which learning results in corticolimbic consolidation of cognitive context models that facilitate new learning in similar contexts.

Sensor density and head surface coverage in EEG source localization

In research with electroencephalographic (EEG) measures, it is useful to identify the sources underlying the potentials recorded at the head surface in order to relate the EEG potentials to brain function. The EEG recorded at the head surface is a function of how current at specific brain (primarily cortical) locations propagates through the conducting volume of head tissues. The accuracy of source localization depends on a sufficient sampling of the surface potential field, an accurate estimation of the conducting volume (head model), and the inverse technique. The present paper reports the effect of spatial sampling of the potential field at the head surface, in terms of both sensor density and coverage of the inferior (lower) as well as superior (upper) head regions. Several inverse methods are examined, using the four shells spherical head model and the finite difference model. Consistent with previous research, greater sensor density improves source localization accuracy. In addition, across all sampling density and inverse methods, sampling across the whole head surface improves the accuracy of source estimates.

An improved artifacts removal method for high dimensional EEG.

BACKGROUND Multiple noncephalic electrical sources superpose with brain signals in the recorded EEG. Blind source separation (BSS) methods such as independent component analysis (ICA) have been shown to separate noncephalic artifacts as unique components. However, robust and objective identification of artifact components remains a challenge in practice. In addition, with high dimensional data, ICA requires a large number of observations for stable solutions. Moreover, using signals from long recordings to provide the large observation set might violate the stationarity assumption of ICA due to signal changes over time. NEW METHOD Instead of decomposing all channels simultaneously, subsets of channels are randomly selected and decomposed with ICA. With reduced dimensionality of the subsets, much less amount of data is required to derive stable components. To characterize each independent component, an artifact relevance index (ARI) is calculated by template matching each component with a model of the artifact. Automatic artifact identification is then implemented based on the statistical distribution of ARI of the numerous components generated. RESULTS The proposed permutation resampling for identification matching (PRIM) method effectively removed eye blink artifacts from both simulated and real EEG. COMPARISON WITH EXISTING METHOD The average topomap correlation coefficient between the cleaned EEG and the ground truth is 0.89±0.01 for PRIM, compared with 0.64±0.05 for conventional ICA based method. The average relative root-mean-square error is 0.40±0.01 for PRIM, compared with 0.66±0.10 for conventional method. CONCLUSIONS The proposed method overcame limitations of conventional ICA based method and succeeded in removing eye blink artifacts automatically.

Neurophysiological Measures of Brain Activity: Going from the Scalp to the Brain

Behavior, such as reaction time and correctness of a response, is the most studied output of the mind in the fields of psychology and human factors. With the advent of modern neuroimaging technologies, opportunities exist for direct study of the minds machinery: the brain. Moreover, there are opportunities for applying these technologies to solve a host of educational and engineering challenges, such as how to design better interfaces with computer systems or how to better educate and train students. The electroencephalogram (EEG) is a direct reflection of the functioning brain, and technologies that enable recording of the EEG have been in existence for more than 50 years. Within the past decade substantial progress has been made in EEG technology, permitting a direct view into the brain. We cover these advances in this paper, which include dense-sensor array technology and physics-based computational head models, and present several examples of how they have been applied.