Neural correlates of error awareness
Introduction
Error processing is necessary for adaptation to changing environmental settings. In order to improve ongoing performance, information about erroneous actions is used to optimize the future response behaviour. While performance monitoring research has attracted great interest, relatively little is known about the role of conscious error perception for action outcome optimization.
Neuroimaging has consistently implicated the posterior frontomedian cortex (pFMC), particularly the rostral cingulate zone (RCZ), in performance monitoring and error processing (Hester et al., 2004, Ullsperger and von Cramon, 2001, Ullsperger and von Cramon, 2004). Event-related brain potential (ERP) research has focused on the error related negativity (ERN), a negative deflection with a fronto-central maximum, peaking 50–100 ms after erroneous responses in choice reaction time tasks (Falkenstein et al., 1990, Gehring et al., 1993). Source localization studies suggest the pFMC to be the main generator of the ERN (Dehaene et al., 1994). This has been corroborated and refined by simultaneous EEG and fMRI recordings revealing a correlation of the single-trial ERN amplitude and the fMRI signal in the RCZ (Debener et al., 2005). Moreover, the single-trial ERN amplitude as well as the fMRI signal in the RCZ predicts post-error slowing, indicating more cautious behaviour after an error had occurred (Debener et al., 2005, Garavan et al., 2002, Kerns et al., 2004). Based on these findings and invasive recordings in human and non-human primates, it has been suggested that the RCZ plays a major role in signalling the need for adjustments in the service of action outcome optimization (Ridderinkhof et al., 2004, Shima and Tanji, 1998, Williams et al., 2004). A second ERP component associated with erroneous responses has been described, the error positivity (Pe; Falkenstein et al., 1990, Falkenstein et al., 2000). It has a centro-parietal maximum and peaks 200–400 ms after the error. The role of the Pe is still rather unclear (Overbeek et al., 2005); it has been suggested to constitute a P3b component associated with the erroneous response. Hajcak et al. (2003) reported a correlation of the Pe amplitude with error-related autonomic responses, specifically skin conductance changes. The neuronal sources of the Pe are still unknown; a source localization study suggests contributions from the pre-genual frontomedian and the parietal cortices (van Veen and Carter, 2002).
Here we investigate how conscious error awareness interacts with the function of the RCZ and post-error adjustments. Previous ERP studies that used an error-signalling procedure in oculomotor tasks have compared errors that were consciously perceived with those which remained unperceived by the subjects. Interestingly, the ERN was present in both aware and unaware errors and was not modulated by conscious awareness (Endrass et al., 2005, Nieuwenhuis et al., 2001). However, both studies revealed a modulation of the Pe: compared to consciously perceived errors, the Pe amplitude was diminished on unaware errors. Interestingly, post-error slowing, i.e., prolonged reaction times subsequent to errors assumed to reflect more cautious responding (Rabbitt, 1966) seems to interact with error awareness. Nieuwenhuis et al. (2001) reported that post-error slowing occurred only on aware errors, while Endrass et al. (2005) found no modulation of post-error slowing by awareness.
A recent neuroimaging study made use of a modified Go/NoGo task with competing response inhibition rules to compare aware and unaware errors (Hester et al., 2005). In accordance with the ERP findings, similar activity in the RCZ was found for aware and unaware errors. Post-error adjustments were modulated by error awareness, showing a typical adjustment pattern after aware errors only.
In addition to cognitive consequences, errors and negative action outcomes are also associated with autonomic reactions. In a number of studies, errors were accompanied by heart rate decelerations, skin conductance changes, and pupilomotor reactions (Critchley et al., 2005b, Crone et al., 2003, Fiehler et al., 2004a, Hajcak et al., 2003, van der Veen et al., 2004). It seems conceivable that error awareness and autonomic responses are related and co-vary. It might be hypothesized that autonomic responses substantially contribute to becoming consciously aware of the error. Thus, interoceptive awareness, reflecting the conscious perception of changes in bodily states, may be closely related to error awareness (Craig, 2002). The anterior inferior insular cortex (located in the vicinity of the polus insulae) has been shown to be involved in visceromotor, i.e., autonomic (Verberne and Owens, 1998) as well as in visceral sensory functions, underlying interoceptive awareness (Critchley, 2005, Critchley et al., 2004). For example, Critchley et al. (2004) showed a clear relationship between the activity of the right anterior insular cortex and the interoceptive accuracy of the subjects in a heartbeat detection task. In another study Critchley et al. (2005a) found a correlation between activity in the left and right insula and heart rate changes evoked by emotional stimuli. Finally, in a numerical version of the Stroop task insular cortex activity was related to errors and sympathetic arousal measured via pupil diameter (Critchley et al., 2005b).
Interestingly, studies on performance monitoring often reported error-related signal increases in a similar insular location (Magno et al., 2006, Ullsperger and von Cramon, 2001, Ullsperger and von Cramon, 2003, Ullsperger and von Cramon, 2004). In summary, it can be hypothesized that activity in the anterior insula is modulated by error awareness.
In the following we report an fMRI study investigating error awareness in an antisaccade task known to yield a roughly equal number of aware and unaware errors (Nieuwenhuis et al., 2001). To inform the region-of-interest (ROI) analysis of our data we carried out an extension of a recent metaanalysis by Ridderinkhof et al. (2004). We employed Activation Likelihood Estimation (ALE; Turkeltaub et al., 2002) to derive a mask for the ROI analysis.
We operationalized error awareness as the ability to explicitly report that a response was erroneous. We tested the question, whether or not error awareness modulates error-related brain activity in the RCZ, the insula or brain areas involved in performance monitoring. Furthermore, the relationship of the brain activity to behavioural adjustments is examined. Does error awareness modulate post-error adjustments and is this reflected in preceding error-related brain activity changes?
Section snippets
Subjects
Thirteen healthy subjects (8 female, mean age = 26.15 years) took part in the study. All participants were right handed and had normal or corrected to normal vision. They reported no neurological or psychological illness at present or in the past and gave written informed consent before the experiment started. Participants agreed to be informed by an inhouse neurologist in case of incidental pathological findings from anatomical scans. The study was performed in accordance with the declaration of
Behavioural data
We discarded trials in which subjects responded too early (reaction time < 80 ms) and trials for which the eye movement data were not interpretable due to technical problems (6.9% ± 1.7 of all trials). Table 2 gives an overview of the behavioural results.
The mean overall error rate across participants was 13.9% (± 1.4). 6.0% (± 0.8) were unaware errors, 7.9% (± 1.0) were aware errors. 64.9% (± 6.6) of the errors were corrected; unaware errors were corrected more often than aware errors (aware, 59.3 ±
Discussion
With the present study we addressed the question, whether or not error awareness modulates error-related brain activity and post-error adjustments. To obtain a roughly equal number of aware and unaware errors we used an antisaccade task (Nieuwenhuis et al., 2001). The brain activity changes we found for error processing confirm the results of previous studies (Ridderinkhof et al., 2004, Ullsperger and von Cramon, 2004): the RCZ, the pre-SMA, and the anterior inferior insular cortex bilaterally
Acknowledgments
The work is supported by grants of the German Research Foundation (Priority Program Executive Functions, SPP 1107) to MU and NK. The help of S. Zysset, M. Naumann, A. Kummer, D. Wilfling, S. Wipper, K. Gille, and B. Schuermann in data acquisition, and of H. Schmidt-Duderstedt for graphical work is greatly acknowledged.
References (44)
- et al.
Activity in the human brain predicting differential heart rate responses to emotional facial expressions
NeuroImage
(2005) - et al.
Anterior cingulate activity during error and autonomic response
NeuroImage
(2005) - et al.
Cardiac concomitants of feedback processing
Biol. Psychol.
(2003) - et al.
ERP components on reaction errors and their functional significance: a tutorial
Biol. Psychol.
(2000) - et al.
Analysis of fMRI time-series revisited
NeuroImage
(1995) - et al.
Event-related fMRI: characterizing differential responses
NeuroImage
(1998) - et al.
Dissociable executive functions in the dynamic control of behavior: inhibition, error detection, and correction
NeuroImage
(2002) - et al.
Neural mechanisms involved in error processing: a comparison of errors made with and without awareness
NeuroImage
(2005) - et al.
LIPSIA—A new software system for the evaluation of functional magnetic resonance images of the human brain
Comput. Med. Imaging Graph.
(2001) - et al.
Characterizing the hemodynamic response: effects of presentation rate, sampling procedure, and the possibility of ordering brain activity based on relative timing
NeuroImage
(2000)
Meta-analysis of the functional neuroanatomy of single-word reading: method and validation
NeuroImage
Subprocesses of performance monitoring: a dissociation of error processing and response competition revealed by event-related fMRI and ERPs
NeuroImage
Neuroimaging of performance monitoring: error detection and beyond
Cortex
Cortical modulation of the cardiovascular system
Prog. Neurobiol.
Analysis of fMRI time-series revisited—Again
NeuroImage
How do you feel? Interoception: the sense of the physiological condition of the body
Nat. Rev. Neurosci.
Neural mechanisms of autonomic, affective, and cognitive integration
J. Comp. Neurol.
Neural activity relating to generation and representation of galvanic skin conductance responses: a functional magnetic resonance imaging study
J. Neurosci.
Neural systems supporting interoceptive awareness
Nat. Neurosci.
Trial-by-trial coupling of concurrent electroencephalogram and functional magnetic resonance imaging identifies the dynamics of performance monitoring
J. Neurosci.
Localization of a neural system for error detection and compensation
Psychol. Sci.
Error awareness in a saccade countermanding task
J. Psychophysiol.
Cited by (325)
Error-related cardiac deceleration: Functional interplay between error-related brain activity and autonomic nervous system in performance monitoring
2024, Neuroscience and Biobehavioral ReviewsSuccess versus failure in cognitive control: Meta-analytic evidence from neuroimaging studies on error processing
2024, Neuroscience and Biobehavioral ReviewsHearts, flowers, and fruits: All children need to reveal their post-error slowing
2023, Journal of Experimental Child Psychology