Neurodynamic Aspects Of Consciousness

Hebb's visionary notion in 1949 of "reverberating cell assemblies" was an important beginning point for a neurodynamic emphasis. Neurodynamics is thus a relatively new discipline, addressing how brain activation changes over time. The neurodynamic perspective is complementary to traditional perspectives that emphasize structures, connectivities, and neuromodulators in that it seeks to understand the time-dependent changes that occur in neuronal populations (neural network models, by comparison, do not reference time). The behavior of these time-sensitive populations are typically measured by EEG, single unit recordings (inso far as these indirectly imply population behaviors), or magnetoencephalography (MEG) and also in dynamic neurochemical measures, such as in vivo dialysis. Neurodynamics attempts to correlate these signatures from various measurement modalities with behavioral and subjective measurements, focusing on the challenge of modeling context-dependent and sequential activation of these highly distributed transient neural ensembles on a moment-to-moment basis.

High levels of temporal resolution are necessary to investigate this, as the neurodynamic integrations that underpin specific "qualia" or subjective content happen quickly but not instantly [requiring, according to some researchers, approximately 300 msec (Libet, 1982)]. Most of the high temporal resolution technologies have poor spatial resolution past the surface of the brain, and even MEG cannot reconstruct neurodynamics in the brainstem.

Without a neurodynamic perspective, neuroscience cannot specify how any physical processes could satisfy the important criteria of isomorphism that many theorists feel is essential to bridging the hard problem: How is it that any aspect of the behavior of neurons can generate phenomenal experience? Most theorists assume that this bridge must be constructed by finding properties of large-scale neuronal ensembles that are functionally isomorphic and temporally coincident with phenomenal experience. Most theorists also agree that neurodynamics must model the integration of top-down and bottom-up processes, and this applies to both early (bottom-up) and late (top-down) sensory cortices, as well as to the larger issue of the relation between brainstem (bottom-up) and cortex (top-down). Neurodynamical models explicating the selectivity (attention), sensory integration, and sense of agency in consciousness would be important bridges indeed.

One of the most puzzling and yet essential properties of consciousness is its seamless integration and fundamental unity. Many investigators have suggested that populations of neurons are coordinated via the generation of coherent patterns or oscillatory envelopes that structure integrative communication between brain regions. Several investigators postulate that the synchronous firing behavior among these distributed populations could constitute the essential neurodynamic underpinnings for conscious states and their contents. Many if not most neurodynamic theories of consciousness are elaborations of basic neuroanatomical concepts that emphasize thalamo-cortical connectivities, and there is relatively little neurodynamic work looking closely at possible contributions of structures underneath the thalamus. These theories propose that essential features of functional integration are achieved thalamocortically, perhaps largely via the functioning and connectivities of the nonspecific thalamic systems (ILN/nRt). However, the lesion correlates that we have summarized in previous sections suggest that these nonspecific thalamic systems are highly dependent upon poorly understood processes in deeper mesodiencephalic regions, as the most severe disorders of consciousness are brought about by damage underneath the thalamus (see case study 3.2).

Singer and Gray et al. (1995) have proposed that neuronal synchronization is necessary for object representation, response selection, sensorimotor integration, and attention. They suggest that temporal synchronization of action potentials in a millisecond range underpins adaptive responses via recruitment of widespread neuronal groups. Corticotectal (not just corticocortical) synchronization has also been found to be crucial, underlining the importance of the various reticular structures just reviewed, with synchronization found to group superior collicular neurons into functionally coherent assemblies. These authors suggest that the stimulus need not come from external sensation, and they do not view the oscillatory activity as a passive response to external stimuli. Instead, they propose that synchronization results at least in part from internally generated goal states, and that external stimulation contributes to the selection of salient goals. Singer et al., argue that the binding that leads to consciousness is brought about by "phase locking" that occurs at single frequencies. From their point of view, the participating neurons sum their trains of action potentials after entrainment. Because the net effect is summed, this model would be considered a linear one, based on essentially proportional relationships. Non linear models do not have this proportion as a crucial feature, as in chaotic systems, where a tiny input can destabilize the system and have profound outcome. Freeman's work (discussed below), highlights the nonlinear, chaotic characteristics of neuronal population behavior.

Edelman's (1987) theory of neuronal group selection is not dissimilar. He has argued that representations arise from a Darwinian-like selection of neuronal groups, that these groups continually interact via "reentry" (reciprocal feedback) and that consciousness emerges from widespread neurodynamic coherences enabled by reentry. Using a 148-channel MEG and a binocular rivalry paradigm, Tononi and Edelman (2000) found that neuronal responses to visual stimuli occurred in a great number of cortical regions, both when the subjects consciously perceived the stimuli and when they did not. However, conscious perception resulted in highly significant differences: "neuromagnetic responses evoked by a stimulus were stronger by 50-85 percent when the subjects were conscious of the stimulus than when they were not conscious" (p. 394). This increase of coherence among various brain regions is consistent with the hypothesis that consciousness reflects rapid integration via reentry.

Both Singer's and Edelman's neurodynamic models advocate a selective and time-dependent coordination of neuronal ensembles that occurs on the order of milliseconds. Both agree upon a nonhierarchical model—that is, there is no reference to layers of binding at different ranges of organizational breadth or complexity. Singer and Edelman state that neuronal groups are selected and reentrantly reselected according to the evolving goals and needs of the organism. Damasio's model, by contrast, suggests that synchrony occurs via a hierarchical effect: Convergence zones located in the association cortices create the binding of lower level neuronal groups.

Llinas and Ribary (1991, 1993), and Joliot et al. (1994) have emphasized the importance of gamma-band 40-Hz oscillations and thalamocortical resonances as essential neurodynamic foundations for consciousness. Llinas and Ribary (1993) studied gamma oscillation in rapid eye movement (REM) sleep and in wakefulness and found coherent 40-Hz activity was evident during REM sleep as well as during wakefulness using a 37-channel MEG. This was the first time that coherent gamma activity was found in REM sleep. No gamma activity was found during delta wave sleep, where consciousness is mostly presumed not to exist. There is evidence that pontine choliner-gic projections into the thalamus are essential for the cortex to organize these fast 40+ Hz oscillatory states and that anticholinergics prevent this, outlining one possible mechanism for the induction of confusional states by anticholinergics (Steriade et al., 1991).

Walter Freeman, a pioneer in neurodynamics, has emphasized the nonlinear, chaotic characteristics of neuronal population behavior, suggesting that the essential neurodynamics of perception and consciousness are nonlinear. Freeman (1975) initially studied population activity via study of the olfactory system, utilizing a 64-lead microelectrode EEG. Freeman proposes that the cortex undergoes global transitions in reaction to meaningful stimuli, settling into a series of neural spatial patterns, a type of oscillatory envelope that is generated by chaotic dynamics of the population. Freeman, who along with Bressler (1980), first coined the term gamma activity to describe this extended synchronous activity, argues that this synchronized gamma activity arises from populations of excitatory and inhibitory neurons in negative feedback. Stimuli drive the system into a transiently more ordered state, in which neural activity can be modeled in terms of mesoscopic wave packets, neural spatial patterns that have time constants closely matching the temporal dynamics of perception and other contents in consciousness (300 to 500 msec). He cites evidence that synchronization is aperiodic and chaotic, spreading across the entire gamma band (40 to 70 Hz), and not reducible to a single frequency (e.g., 40 Hz). Consciousness, according to Freeman, emerges from how such spatially and temporally extended neural patterns across the gamma band underwrite widespread integration of brain activity. This kind of brain activity enables not a static set of representations but a highly plastic and evolving system of meanings for the organism.

0 0

Post a comment