Temporal unpredictability increases error monitoring as revealed by EEG-EMG investigation.

Reacting in an unpredictable context increases error monitoring as evidenced by greater error-related negativity (ERN), an electrophysiological marker linked to an evaluation of response outcomes. We investigated whether ERN also increased when participants evaluated their responses to events that appeared in unpredictable versus predictable moments in time. We complemented electroencephalographic (EEG) analysis of cortical activity by measuring performance monitoring processes at the peripheral level using electromyography (EMG). Specifically, we used EMG data to quantify how temporal unpredictability would affect motor time (MT), the interval between the onset of muscle activity, and the mechanical response. MT increases following errors, indexing online error detection, and an attempt to stop incorrect actions. In our temporally cued version of the stop-signal task, symbolic cues predicted (temporally predictable condition) or not (temporally unpredictable condition) the onset of a target. In 25% of trials, an auditory signal occurred shortly after the target presentation, informing participants that they should inhibit their response completely. Response times were slower, and fewer inhibitory errors were made during temporally unpredictable than predictable trials, indicating enhanced control of unwanted actions when target onset time was unknown. Importantly, the ERN to inhibitory errors was greater in temporally unpredictable relative to temporally predictable conditions. Similarly, EMG data revealed prolonged MT when reactions to temporally unpredictable targets had not been stopped. Taken together, our results show that a temporally unpredictable environment increases the control of unwanted actions, both at cortical and peripheral levels, suggesting a higher subjective cost of maladaptive responses to temporally uncertain events.

Deciphering interference control in adults with ADHD by using distribution analyses and electromyographic activity.

A deficit in "interference control" is commonly found in adults with Attention Deficit Hyperactivity Disorder (ADHD). This has mainly been interpreted as difficulties in inhibiting inappropriate responses. However, interference control involves processes other than simply the ability to inhibit. Consequently, we used sophisticated analysis to decipher the additional processes of interference control in these patients. We compared interference control between 16 adults with ADHD and 15 control adults performing a Simon task. In most studies, performance is generally reported in terms of mean error rates and reaction times (RTs). However, here we used distribution analyses of behavioral data, complemented by analyses of electromyographic (EMG) activity. This allowed us to better quantify the control of interference, specifically the part that remains hidden when pure correct trials are not distinguished from partial errors. Partial errors correspond to sub-threshold EMG bursts induced by incorrect responses that immediately precede a correct response. Moreover, besides "online" control, we also investigated cognitive control effects manifesting across consecutive trials. The main findings were that adults with ADHD were slower and showed a larger interference effect in comparison to controls. However, the data revealed that the larger interference effect was due neither to higher impulse expression, nor to a deficit in inhibition but that these patients presented a larger interference effect than the controls after congruent trials. We propose and discuss the hypothesis that the interference control deficit found in adults with ADHD is secondary to impairments in sustained attention.

Detecting single-trial EEG evoked potential using a wavelet domain linear mixed model: application to error potentials classification.

OBJECTIVE: The main goal of this work is to develop a model for multisensor signals, such as magnetoencephalography or electroencephalography (EEG) signals that account for inter-trial variability, suitable for corresponding binary classification problems. An important constraint is that the model be simple enough to handle small size and unbalanced datasets, as often encountered in BCI-type experiments. APPROACH: The method involves the linear mixed effects statistical model, wavelet transform, and spatial filtering, and aims at the characterization of localized discriminant features in multisensor signals. After discrete wavelet transform and spatial filtering, a projection onto the relevant wavelet and spatial channels subspaces is used for dimension reduction. The projected signals are then decomposed as the sum of a signal of interest (i.e., discriminant) and background noise, using a very simple Gaussian linear mixed model. MAIN RESULTS: Thanks to the simplicity of the model, the corresponding parameter estimation problem is simplified. Robust estimates of class-covariance matrices are obtained from small sample sizes and an effective Bayes plug-in classifier is derived. The approach is applied to the detection of error potentials in multichannel EEG data in a very unbalanced situation (detection of rare events). Classification results prove the relevance of the proposed approach in such a context. SIGNIFICANCE: The combination of the linear mixed model, wavelet transform and spatial filtering for EEG classification is, to the best of our knowledge, an original approach, which is proven to be effective. This paper improves upon earlier results on similar problems, and the three main ingredients all play an important role.

Detecting and correcting partial errors: Evidence for efficient control without conscious access.

Appropriate reactions to erroneous actions are essential to keeping behavior adaptive. Erring, however, is not an all-or-none process: electromyographic (EMG) recordings of the responding muscles have revealed that covert incorrect response activations (termed "partial errors") occur on a proportion of overtly correct trials. The occurrence of such "partial errors" shows that incorrect response activations could be corrected online, before turning into overt errors. In the present study, we showed that, unlike overt errors, such "partial errors" are poorly consciously detected by participants, who could report only one third of their partial errors. Two parameters of the partial errors were found to predict detection: the surface of the incorrect EMG burst (larger for detected) and the correction time (between the incorrect and correct EMG onsets; longer for detected). These two parameters provided independent information. The correct(ive) responses associated with detected partial errors were larger than the "pure-correct" ones, and this increase was likely a consequence, rather than a cause, of the detection. The respective impacts of the two parameters predicting detection (incorrect surface and correction time), along with the underlying physiological processes subtending partial-error detection, are discussed.

Simon effect in the rat: a new model for studying the neural bases of the dual-route architecture.

In humans, the Simon effect refers to the fact that choice reaction time (RT) is shorter when the stimulus corresponds spatially to the response than when it does not, albeit the location of the stimulus is irrelevant to the task. This effect has motivated innumerable empirical and theoretical studies and is considered to reflect elementary cognitive processes. We report an experiment demonstrating that rats also display a Simon effect, the dynamics of which--as assessed by factorial manipulations and RT distribution analyses--partly corresponds to those of the effect studied in human participants. The present results are consistent with the ideas that in rats, like in humans, (i) the information conveyed by the stimulus is processed via two parallel routes, one controlled and relatively slow, and one fast and automatic (dual-route architecture) and (ii) the dual-route processing is finished before the start of motor processes. The correspondence between these findings and those reported in humans open new perspectives for neurophysiological investigations of the dual-route architecture in an animal model routinely studied in neuroscience research.

Spatial enhancement of EEG traces by surface Laplacian estimation: comparison between local and global methods.

OBJECTIVE: Surface Laplacian estimation enhances EEG spatial resolution. In this paper, we compare, on empirical grounds, two computationally different estimations of the surface Laplacian. METHODS: Surface Laplacian was estimated from the same monopolar data set with both Hjorth's method [local; Electroenceph Clin Neurophysiol 39 (1975) 526] as modified by MacKay [Electroenceph Clin Neurophysiol 56 (1983) 696] and with spherical spline interpolation [global; Electroenceph Clin Neurophysiol 72 (1989) 184]. RESULTS: The grand averages computed with the two methods proved to be very similar but differed markedly from the monopolar ones. The two different computations were highly correlated, presented low relative errors and allowed to evidence comparable experimental effects. CONCLUSIONS: These results suggest that Hjorth's method and spherical spline interpolation convey similar topographic and chronometric informations. SIGNIFICANCE: We provide empirical evidence that local and global methods of surface Laplacian estimation are equivalent to improve the spatial resolution of EEG traces. Global methods allow to explore the scalp topography and local methods allow to spare time in electrode setting that can be useful for studies on special populations (i.e. children, aged subjects) and for clinical purposes.

Response monitoring without sensory feedback.

OBJECTIVE: The elicitation of an evoked potential, the 'error negativity' (Ne) when subjects commit errors in speeded tasks, is often taken as an index of response monitoring processes. The presence of a Ne-like wave on purely correct trials challenges the current conceptions about the nature of such a monitoring system. Here, we evaluate the possibility that the Ne-like wave on correct trials is merely due to reafferences, and at the same time, we test directly the general opinion according to which the Ne is generated by an internal signal. METHODS: We studied the presence of a Ne-like wave in a completely deafferented patient. The patient performed two reaction time (RT) tasks: a two-responses RT task and a go/no-go task. RESULTS: In this patient, a Ne occurs on errors, on incorrect EMG activations, and on purely correct responses. On errors, the Ne was clearly followed by an error positivity (Pe). CONCLUSIONS: The Ne and the Ne-like wave are not generated by reafferences. This similarity is a further argument to consider that these two waves are of same nature. SIGNIFICANCE: The present data demonstrate that sensory information is not mandatory for the brain to monitor and correct ongoing responses.

The influence of time preparation on motor processes assessed by surface Laplacian estimation.

OBJECTIVE: The present study was aimed at testing whether foreperiod duration affects the activity recorded over the primary sensorimotor cortices during the reaction time. METHODS: The foreperiod duration (500 or 2500 ms) was varied across blocks of trials during a between-hand choice reaction time task; surface Laplacians were estimated from EEG recordings by the source derivation method. RESULTS: Reaction time was shorter for the 500 ms foreperiod than for the 2500 ms foreperiod. A contralateral negativity/ipsilateral positivity pattern showed up over the primary sensorimotor cortices. The time between the contralateral negativity onset and the electromyographic onset was shorter for the 500 ms foreperiod than for the 2500 ms foreperiod, which suggests that the foreperiod affects the implementation of the motor command. Furthermore, the interval between the onset of the voluntary electromyographic activity and the mechanical response was shorter for the 500 ms foreperiod than for the 2500 ms foreperiod. CONCLUSIONS: These results indicate that time preparation affects both central and peripheral motor processes.

Error negativity on correct trials: a reexamination of available data.

The error negativity, an EEG wave observed when subjects commit an error in a choice reaction time (RT) task, is often considered as a sign of error detection. Recently, reports of Ne-like waves on correct responses did challenge this interpretation. It has been proposed, however, that these Ne-like waves result either from an artifactual contamination of response-locked activities by stimulus-locked ones, or from an implicit monitoring of the time elapsing during the RT. Our aim was to reprocess published data: (1) to compare the shape and amplitude of EMG-locked and stimulus-locked ERPs on correct trials, and (2) to compare the size of the EMG-locked Ne-like waves obtained on fast and slow trials. The results neither support the artifact hypothesis nor the RT monitoring one. Therefore, it seems that the Ne-like waves observed on correct trials do correspond to a Ne, which suggests that the Ne has a broader significance than just error detection.

Deficit in motor cortical activity for simultaneous bimanual responses.

Reaction time (RT) is known to be longer for simultaneous bimanual responses than for unimanual ones. This phenomenon is called "bilateral deficit". To identify the mechanisms subserving the bilateral deficit, brain electrical activity was examined, with a source derivation method, in 12 right-handed subjects, during the preparation and execution periods of a RT task. The responses were either unilateral or bilateral index finger flexion, performed either in a simple RT condition, with 20% catch trials, or in a choice RT condition. A deficit was observed in RT for the bilateral response for the right-index finger movement. In cerebral electrical activities, no evidence of a correlate of a bilateral deficit was found during the preparatory period. Conversely, during the execution period, an EEG correlate of the bilateral deficit was found. For the right hand, the activation of the sensorimotor area directly involved in the voluntary control was weaker for bilateral than for unilateral contralateral responses. The reasons for such a bilateral command weakness are discussed in the context of our RT task. First, the constraint of synchronisation included in the bilateral response might require an interhemispheric information transmission that resulted in a braking effect. Second, given that an ipsilateral inhibition is present in case of choice between the two hands of one particular unimanual response, and given that this ipsilateral inhibition is also present in case of simple unimanual trials, we hypothesise that a mutual transcallosal inhibitory effect also persists in the bilateral response.

An electromyographic investigation of the effect of stimulus-response mapping on choice reaction time.

The activity of the agonist muscles was recorded during the performance of a two-choice visual reaction time (RT) task in which the compatibility of the stimulus-response mapping was manipulated. Correct trials were distinguished according to whether or not the activation of the agonist of the required response was preceded by an activation of the agonist of the nonrequired response. Double activation trials were more numerous for the incompatible than for the compatible mapping. Furthermore, these trials yielded longer RTs than the single muscular activation trials. These results suggest that initial activations of nonrequired responses are more frequently aborted and corrected when the mapping is incompatible than when it is compatible. This finding supports the dimensional overlap model of stimulus-response compatibility (S. Kornblum, T. Hasbroucq, & A. Osman, 1990).

Dissociation between activation and attention effects in time estimation: implications for internal clock models.

In a time production task, the participants' activation level and attention devoted to time were manipulated respectively by means of click trains delivered at 2 different intensities during the task and by introducing a concurrent reaction time task. Activation level is classically considered to affect the rate of an internal pacemaker, whereas the way attention affects time estimation is a matter of debate. Three models that differ as to the effect of attention were evaluated. Predictions on the interaction pattern between activation and attention were derived for each of the 3 models. When manipulated jointly, these 2 factors proved to be independent, as they had additive effects on the performance. This finding suggests that the activation level affects the pacemaker rate, whereas the attention level affects an accumulation process by directly acting on a switch functioning in an all-or-none fashion.

Changes in spinal excitability during choice reaction time: the H reflex as a probe of information transmission.

The aim of the present study was to investigate the modulations in amplitude of H reflexes elicited in a hand muscle, the flexor pollicis brevis, during the performance of a choice reaction time (RT) task in which this muscle was directly involved. Ten subjects were to choose between a left- or a right-thumb key-press according to the lateral location of a flash of light. The stimulus-response mapping was either compatible or incompatible. Hoffman reflexes were elicited at different times during the RT by stimulation of the median nerve. Twenty-five milliseconds before the voluntary response, the amplitude of the H reflex suddenly increased when the muscle was involved in the response and decreased symmetrically when the muscle was not involved in the response. Mapping compatibility exerted no detectable influence on the changes in spinal excitability. The latter result supports the assumptions that are at the core of Sternberg's additive factor method.

High-speed memory scanning: a behavioral argument for a serial oscillatory model.

In order to account for the memory span [G.A. Miller, The magical number seven, plus minus two: some limits on our capacity for processing information, Psychol. Rev. 63 (1956) 81-97.], the magical number seven, plus minus two, and high-speed scanning in human memory ¿S. Sternberg, High speed scanning in human memory, Science 153 (1966) 652-654., Lisman and collaborators [O. Jensen, J.E. Lisman, An oscillatory short-term memory buffer model can account for data on the Sternberg task, J. Neurosci. 18 (1998) 10688-10699; J.E. Lisman, M.A.P. Idiart, Storage of 7+/-2 short-term memories in oscillatory subcycles, Science 267 (1995), 1512-1515.] proposed an oscillatory short-term memory buffer model. In this neurophysiological model: "a single brain network can separately maintain up to seven memories by a multiplexing mechanism that uses theta and gamma brain oscillations for clocking. A memory is represented by groups of neurons that fire in the same gamma cycle" ¿O. Jensen, J.E. Lisman, An oscillatory short-term memory buffer model can account for data on the Sternberg task, J. Neurosci. 18 (1998) 10688-10699, p. 10688. To test this model, we tried to modify the memory scanning time by shifting the gamma oscillation frequency. To this aim, we replicated the visual short-term memory scanning task ¿S. Sternberg, High speed scanning in human memory, Science 153 (1966) 652-654., and we simultaneously used the protocol that Treisman ¿M. Treisman, A. Faulkner, P.L.N. Naish, D. Brogan, The internal clock: evidence for a temporal oscillator underlying time perception with some estimates of its characteristics frequency, Perception 19 (1990) 705-743. designed to drive, slowing down or speeding up, a temporal oscillator acting in the gamma range ¿J.G.R. Jefferys, R.D. Traub, M.A. Whittington, Neuronal networks for induced "40 Hz rhythms, Trends Neurosci. 19 (1996) 202-208; W. MacKay, Synchronized neuronal oscillations and their role in motor processes, Trends Cog. Sci. 1 (1997) 176-183; M. Treisman, N. Cook, P.L.N. Naish, J.K. MacCrone, The internal clock: electroencephalographic evidence for oscillatory processes underlying time perception, Q. J. Exp. Psychol. 47A (1994) 241-289.. In this protocol, an auditory periodic stimulus (click train) was delivered at various frequencies during the task. The reaction time (RT), the slope, and the intercept of the linear function associating RT to memorized list length showed systematic modulations according to the stimulation frequency. The predicted driving effects due to the click trains were obtained, consisting of localised modulations of performance on the stimulation frequency band. We argue that memory scanning is indeed paced by a temporal oscillator, thus providing behavioral arguments for the serial oscillatory model of Lisman.

Cortico-spinal inhibition reflects time but not event preparation: neural mechanisms of preparation dissociated by transcranial magnetic stimulation.

Changes in cortico-spinal excitability related to time and event preparation were investigated by transcranial magnetic stimulation (TMS) of the motor cortex during the foreperiod of a movement-precuing task. Subjects performed a four alternative choice reaction time (RT) task involving a button-press with the index or middle finger (FI) of the left or right hand. Advance information about the to-be-signaled response was provided by a precue, which preceded the response signal by a 1 s foreperiod. The precue either indicated the hand (right or left) or FI (index or middle) with which the response would be executed or was uninformative. TMS was delivered to the left or right cortical hand area at one of five possible times during the foreperiod: -1000, -500, -333, -166 or 0 ms prior to the response signal. Surface EMG activity from a prime mover involved in flexion of the response FIs (Flexor digitorum superficialis) was used to measure the magnitude of the motor evoked potential (MEP) elicited by TMS. Cortico-spinal excitability--as assessed by the magnitude of the MEP evoked in the target muscle contralateral to the stimulated hemisphere--progressively decreased during the foreperiod. The identity of the precued responses, however, had no effect on MEP magnitude. These results suggest that preparation to respond at a particular time inhibited excitability of the cortico-spinal tract, while advance preparation to perform specific responses affected more central structures only.

What's an internal clock for? From temporal information processing to temporal processing of information.

The existence of an internal clock and its involvement in information processing has been investigated in humans using the experimental protocol of Treisman et al. (1990) [Treisman et al., 1990. Perception, 19, 705-743]. In this protocol, a periodical stimulation, which is assumed to drive an internal clock, is delivered during a reaction time (RT) task. The accelerating or slowing down effects of the periodical stimulation, according to its frequency, allowed an estimate to be made of a simple harmonic of the frequency of the internal clock. The estimate was close to 21 Hz. In the framework of the serial model of information processing, the present work investigates the involvement of the internal clock in the transmission of information between processing stages during RT. The data tend to support the idea that the internal clock allows the transfer of information from one stage to the next one at definite moments only, periodically distributed in time. According to our results, and recent data from the literature on electric cortical oscillations, we propose a model of an internal clock sending periodic inhibition, which would permit an increased signal/noise ratio in the processing and the transmission of information in the central nervous system.

Further argument for the existence of a pacemaker in the human information processing system.

To support the idea that temporal information processing may depend on an internal clock, Treisman et al. proposed a pacemaker model (Treisman, M., Faulkner, A., Naish, P.L.N., Brogan, D., 1990. The internal clock: Evidence for a temporal oscillator underlying time perception with some estimates of its characteristics frequency. Perception 19, 705-743.) and a technique for interfering with it by introducing an external periodic phenomenon. Experimental results obtained by these authors on time estimation and production tasks support this model. In another study, Treisman et al. established that the pacemaker also affects reaction times (RT) (Treisman, M., Faulkner, A., Naish, P.L.N., 1992. On the relation between time perception and the timing of motor action: Evidence for a temporal oscillator controlling the timing of movement. Quarterly Journal of Experimental Psychology 45A, 235-263.). In the present study, we addressed the question as to which information processing stage (Sanders, A.F., 1980. Stage analysis of reaction process, In: Stelmach, G.E., Requin, J. (Eds.). Tutorials in motor behavior. North-Holland, Amsterdam, pp. 331-354.) is affected by this internal clock. For this purpose, we used the Additive Factors Method (Sternberg, S., 1969. The discovery of processing stages: Extension of Donder's method. In: Koster, W.G. (Ed.). Attention and Performance II. Acta Psychologica 30, 276-315.). To vary sensorial processing time, we used two visual stimulus intensities. Stimulus-response mapping was manipulated to enhance central processing time. To modify the duration of the motor stages, the two responses could be given by two fingers on the same hand (right ring vs. middle finger) or by two fingers of the different hands (right ring vs. left middle finger). Intensity of the stimulus, stimulus-response mapping, and repertoire of responses were found to be additive. We obtained RT modulations similar to those obtained by Treisman et al. in 1992. No first order interactions were observed between the periodical phenomenon and the other manipulated factors but only a third order one. Two possible interpretations of these results are proposed.