When the brain changes its mind: flexibility of action selection in instructed and free choices

The neural mechanisms underlying the selection and initiation of voluntary actions in the absence of external instructions are poorly understood. These mechanisms are usually investigated using a paradigm where different movement choices are self-generated by a participant on each trial. These "free choices" are compared with "instructed choices," in which a stimulus informs subjects which action to make on each trial. Here, we introduce a novel paradigm to investigate these modes of action selection, by measuring brain processes evoked by an instruction to either reverse or maintain free and instructed choices in the period before a "go" signal. An unpredictable instruction to change a response plan had different effects on free and instructed choices. In instructed trials, change cues evoked a larger P300 than no-change cues, leading to a significant interaction of choice and change condition. Free-choice trials displayed a trend toward the opposite pattern. These results suggest a difference between updating of free and instructed action choices. We propose a theoretical framework for internally generated action in which representations of alternative actions remain available until a late stage in motor preparation. This framework emphasizes the high modifiability of voluntary action.

Dissociations between motor-related EEG measures in a cued movement sequence task

Different aspects of preparation, especially processes related to knowing what to prepare versus applying that foreknowledge effectively, may be reflected in different types of brain activity, e.g., the lateralized readiness potential (LRP), beta-band event-related desynchronization and phase locking. In a previous study in which subjects had to switch between response hands, dissociations were found between types of preparatory hand-related lateralization in evoked potentials, amplitude and phase locking (Gladwin et al., 2006) knowing what task set to switch to and effectively preparing that task set affected the different measures of lateralization. Similarly to how, in task switching, stimuli and responses must be correctly related to each other, in the preparation of movement sequences relations must be specified concerning motor processes. Similar dissociations as found in the task switching data might then be found in a cued movement sequence task. This possibility was explored by precueing elements of a two-movement sequence involving the left and right index fingers, and comparing various measures of electroencephalogram activity. Cues could specify the full sequence, either the first or the second element, or neither element. Knowing the first element was sufficient to lateralize the pattern of phase locking, but effects were found in the LRP and lateralized amplitude only when the full sequence was known. It seems likely that subjects only fully prepared the first response when they had full knowledge of the sequence so that the dissociation may be closely related to that found for task switching. Thus, the present data would appear to agree with previous results that couple response-foreknowledge with phase locking and the transformation of that foreknowledge into effective changes in component processes with evoked potential shifts. The results further underscore the general importance of considering different types of brain activity: depolarization, desynchronization and phase locking all appear to be involved in different aspects of cognitive control.

On the relationship between slow cortical potentials and BOLD signal changes in humans

This review summarizes experimental studies that investigated the relationship between DC-recorded slow event-related potentials (slow waves) of the electroencephalogram (EEG) and the hemodynamic BOLD response, as measured with functional magnetic resonance imaging (fMRI). Slow waves have been found to accompany a large number of cognitive processes in a systematic and topographically specific way, and have thus been successfully employed in psychophysiological experiments to dissociate cognitive functions by means of their slow wave topography. Recently, however, several independent studies, using different experimental paradigms, suggest the existence of another feature of slow waves, i.e., a close relationship with the fMRI BOLD response. Some of these studies found couplings between slow waves and BOLD signals in various brain regions, using simultaneous EEG-fMRI recordings. Others found similar task-related activation patterns of slow waves (i.e., scalp topographies) and BOLD responses (i.e., activated voxel profiles), as well as corresponding parametric increases of signal strength with increasing task difficulty. The close relationship between slow waves and BOLD responses reported here concerns a low frequency range of the EEG signal (<1 Hz) that has so far received less attention in studies on the correspondence between EEG and fMRI than the higher frequencies, and therefore adds to the various findings obtained at higher EEG frequencies. Implications for the use of slow waves for neuroscientific research are discussed.

Tracing sequential waves of rapid visuomotor activation in lateralized readiness potentials

Feedforward activation processes are widely regarded as crucial for the automatic initiation of motor responses, whereas recurrent processes are often regarded as crucial for visual awareness. Here, we used a set of behavioral criteria to evaluate whether rapid processing in the human visuomotor system proceeds as would be expected of a feedforward system that works independent of visual awareness. We measured lateralized readiness potentials (LRPs) for key-press responses to color targets ("masks") preceded by masked color primes mapped onto the same or opposite response, and traced the time-course of motor activation as a function of different prime and mask characteristics. LRP time-courses showed that initial motor activation occurred in prime direction and was time-locked to prime onset. Response activation was then captured on-line by the mask signal, with motor activation now time-locked to the mask and proceeding in mask direction. Crucially, the time-course of early activation by the prime was independent of all mask characteristics. This invariance in early priming effects contrasted with large differences in visual awareness for the prime produced by different masks. Results suggest that primed responses to color stimuli are controlled by feedforward waves of activation sequentially elicited by prime and mask signals traveling rapidly enough to escape the recurrent processes leading to backward masking.

Brain potentials indicate poor preparation for action in schizophrenia

Impaired antisaccade performance in schizophrenia (SZ) may originate from poor task preparation, suggested by low amplitudes of the contingent negative variation (CNV) before antisaccades. To dissociate components of preparation we measured the CNV in standard pro- and antisaccades and a stimulus preceding negativity (SPN) in delayed pro- and antisaccades. In healthy controls the SPN had lower amplitudes than the CNV, reflecting mere stimulus expectation in SPN and combined stimulus expectation and action readiness in CNV. SZ patients had lower CNV amplitudes than controls, but there was no reliable indication of a general SPN reduction, suggesting a particular impairment of action readiness. The CNV and SPN amplitudes of controls were larger if tasks had incongruent (anti) compared to congruent (pro) S-R mappings. This difference was absent in SZ patients, suggesting a failure to activate specific resources for incongruent S-R mappings.

Effects of acoustic startle stimuli on interceptive action

In reaction time (RT) tasks, presentation of a startling acoustic stimulus (SAS) together with a visual imperative stimulus can dramatically reduce RT while leaving response execution unchanged. It has been suggested that a prepared motor response program is triggered early by the SAS but is not otherwise affected. Movements aimed at intercepting moving targets are usually considered to be similarly governed by a prepared program. This program is triggered when visual stimulus information about the time to arrival of the moving target reaches a specific criterion. We investigated whether a SAS could also trigger such a movement. Human experimental participants were trained to hit moving targets with movements of a specific duration. This permitted an estimate of when movement would begin (expected onset time). Startling and sub-startle threshold acoustic probe stimuli were delivered unexpectedly among control trials: 65, 85, 115 and 135 ms prior to expected onset (10:1 ratio of control to probe trials). Results showed that startling probe stimuli at 85 and 115 ms produced early response onsets but not those at 65 or 135 ms. Sub-threshold stimuli at 115 and 135 ms also produced early onsets. Startle probes led to an increased vigor in the response, but sub-threshold probes had no detectable effects. These data can be explained by a simple model in which preparatory, response-related activation builds up in the circuits responsible for generating motor commands in anticipation of the GO command. If early triggering by the acoustic probes is the mechanism underlying the findings, then the data support the hypothesis that rapid interceptions are governed by a motor program.

Very high frequency oscillations (VHFO) as a predictor of movement intentions

Gamma band (30-80 Hz) oscillations arising in neuronal ensembles are thought to be a crucial component of the neural code. Recent studies in animals suggest a similar functional role for very high frequency oscillations (VHFO) in the range 80-200 Hz. Since some intracerebral studies in humans link VHFO to epileptogenesis, it remains unclear if VHFO appear in the healthy human brain and if so which is their role. This study uses EEG recordings from twelve healthy volunteers, engaged in a visuo-motor reaction time task, to show that VHFO are not necessarily pathological but rather code information about upcoming movements. Oscillations within the range (30-200 Hz) occurring in the period between stimuli presentation and the fastest hand responses allow highly accurate (>96%) prediction of the laterality of the responding hand in single trials. Our results suggest that VHFO belong in functional terms to the gamma band that must be considerably enlarged to better understand the role of oscillatory activity in brain functioning. This study has therefore important implications for the recording and analysis of electrophysiological data in normal subjects and patients.

Pre-stimulus EEG effects related to response speed, task switching and upcoming response hand

The task-switching paradigm provides an opportunity to study whether oscillatory relations in neuronal activity are involved in switching between and maintaining task sets. The EEG of subjects performing an alternating runs [Rogers, R.D., Monsell, S., 1995. Costs of a predictable switch between simple cognitive tasks. Journal of Experimental Psychology: General 124, 207-231] task-switching task was analyzed using event-related potentials, the lateralized readiness potential, instantaneous amplitude and the phase-locking value [Lachaux, J.P., Rodriguez, E., Martinirie, J., Varela, F.J., 1999. Measuring phase synchrony in brain signals. Human Brain Mapping 8, 194-208]. The two tasks differed in the relevant modality (visual versus auditory) and the hand with which responses were to be given. The mixture model [de Jong, R., 2000. An intention driven account of residual switch costs. In: Monsell, S., Driver, J. (Eds.), Attention and Performance XVII: Cognitive Control. MIT Press, Cambridge] was used to assign pre-stimulus switch probabilities to switch trials based on reaction time; these probabilities were used to create a fast-slow distinction between trials on both switch and hold trials. Results showed both time- and time-frequency-domain effects, during the intervals preceding stimuli, of switching versus maintenance, response speed of the upcoming stimulus, and response hand. Of potential importance for task-switching theory were interactions between reaction time by switch-hold trial type that were found for a frontal slow negative potential and the lateralized readiness potential during the response-stimulus interval, indicating that effective preparation for switch trials involves different anticipatory activity than for hold trials. Theta-band oscillatory activity during the pre-stimulus period was found to be higher when subsequent reaction times were shorter, but this response speed effect did not interact with trial type. The response hand of the upcoming task was associated with lateralization of pre-stimulus mu- and beta-band amplitude and, specifically for switch trials, beta-band phase locking.

Semantic activation in action planning

Four experiments investigated activation of semantic information in action preparation. Participants either prepared to grasp and use an object (e.g., to drink from a cup) or to lift a finger in association with the object's position following a go/no-go lexical-decision task. Word stimuli were consistent to the action goals of the object use (Experiment 1) or to the finger lifting (Experiment 2). Movement onset times yielded a double dissociation of consistency effects between action preparation and word processing. This effect was also present for semantic categorizations (Experiment 3), but disappeared when introducing a letter identification task (Experiment 4). In sum, our findings indicate that action semantics are activated selectively in accordance with the specific action intention of an actor.