Core networks for visual-concrete and abstract thought content: A brain electric microstate analysis

Abstract

Commonality of activation of spontaneously forming and stimulus-induced mental representations is an often made but rarely tested assumption in neuroscience. In a conjunction analysis of two earlier studies, brain electric activity during visual-concrete and abstract thoughts was studied. The conditions were: in study 1, spontaneous stimulus-independent thinking (post-hoc, visual imagery or abstract thought were identified); in study 2, reading of single nouns ranking high or low on a visual imagery scale. In both studies, subjects' tasks were similar: when prompted, they had to recall the last thought (study 1) or the last word (study 2). In both studies, subjects had no instruction to classify or to visually imagine their thoughts, and accordingly were not aware of the studies' aim. Brain electric data were analyzed into functional topographic brain images (using LORETA) of the last microstate before the prompt (study 1) and of the word-type discriminating event-related microstate after word onset (study 2). Conjunction analysis across the two studies yielded commonality of activation of core networks for abstract thought content in left anterior superior regions, and for visual-concrete thought content in right temporal-posterior inferior regions. The results suggest that two different core networks are automatedly activated when abstract or visual-concrete information, respectively, enters working memory, without a subject task or instruction about the two classes of information, and regardless of internal or external origin, and of input modality. These core machineries of working memory thus are invariant to source or modality of input when treating the two types of information.

Introduction

A principal feature of brain functioning is that incoming information activates content-specific neural networks. Since the brain is able to maintain a representation of the incoming information beyond the time when the receptors were active, these networks must be able to maintain their state by themselves for some period of time. One could see this as a notion of working memory in the basic view of computational neuroscience.

In EEG potentials time-locked to incoming information (Event-Related Potentials, ‘ERP’), sequences of brief time periods of quasi-stable brain electric field topography are observed (‘microstates’) that are concatenated by rapid changes of topography (Lehmann and Skrandies, 1980, Brandeis et al., 1995, Michel et al., 2001, Lehmann et al., 2009). Since different potential landscapes must have been produced by differently active neural populations—wherever in the brain they might be—, it is reasonable to assume that different microstates perform different functions. Accordingly, periods of stable topography likely indicate periods of stable brain functional state. Sequences of microstates may thus reflect a sequential activation of extended neuro-cognitive networks, each representing specific aspects of the incoming information.

Microstates were also observed in so-called ‘spontaneous’ EEG (Lehmann, 1971, Lehmann et al., 1987, Lehmann et al., 2005). The duration of these temporal brain electric microstates in spontaneous EEG is well in the subsecond range, from about 70 to 150 ms, thus qualifying for the postulated constituents of the seemingly continual ‘stream of consciousness’ (James, 1890), for the building blocks of mentation or ‘atoms of thought’ (Lehmann et al., 1998).

In a no-task condition with closed eyes, i.e. in a day-dream situation, subjects experience a flow of mentation, the ‘stream of consciousness’. When this stream of mentation is intercepted at random by gentle prompt signals that ask the subjects to verbally report their last experience, most of such reports can be easily sorted into two classes, mental imagery and abstract thoughts. For these two classes of experience, multichannel brain electric field recordings (EEG) significantly differed in spatial distribution of the electric potential during the microstate that happened to be present during the prompts (Lehmann et al., 1998). Thus, different brain networks were active for visual and abstract mentation.

In an experiment where subjects read words for later recall without a task to imagine or to classify the words, an ERP microstate was observed that had a significantly different spatial distribution of the brain electric field for words that were rated high versus words that were rated low on a visual imageability scale. This microstate lasted from 286 to 354 ms after stimulus onset (Koenig et al., 1998). The onset time of the distinguishing microstate at 286 ms after stimulus onset was very close to those in recent ERP and MEG studies on reading concrete and abstract words, both asking for task executions (Dhond et al. 2007: 330 ms; Wirth et al. 2008: 280 ms).

Thus, both our EEG and our ERP studies had shown a difference in microstate potential map landscape between the visual and abstract conditions, even though the two studies used a completely different experimental design, i.e. spontaneous thoughts versus word reading. On the other hand, there was commonality of their results (Fig. 1): in both studies, the electric axis of the microstate field for the visual condition was rotated counter-clockwise compared to that of the abstract condition; in addition, in both studies, the location of the brain electric gravity center of the microstate field was more to the right for the visual than for the abstract condition (Koenig et al., 1998, Lehmann et al., 1998).

The question arises whether our data is consistent with the hypothesis that there are specific core regions that are automatedly activated during processing of visual-concrete and abstract information, independent of the evoking condition (internally generated or externally input-driven), and independent of a task that required the subject to willfully generate such states. This question is interesting also in the light of recent fMRI studies showing, albeit on a much slower time scale (Mantini et al., 2007), that resting state networks observed during no-task-conditions resembled networks activated during task executions (Damoiseaux et al., 2006, De Luca et al., 2006).

Accordingly, we tested the hypothesis that common in the two experiments, the visual experience would show significant activation in particular brain regions, and the abstract experience in other particular regions. The regions were identified using Low Resolution Electromagnetic tomography (LORETA, Pascual-Marqui et al., 1994, Pascual-Marqui et al., 1999) that computes from the measured potential distributions on the head surface the localization of the intracerebral, cortical generators in a three-dimensional voxel space.