National Science Centre OPUS Research Project (2011-2014)

"Neural systems for error monitoring: activity patterns of detection and correction of erroneous saccadic reactions and their circadian stability in chronic sleep deficit state - simultaneous dense array EEG and oculographic study"

National Science Centre Research Project registered as 2011/01/B/HS6/00446

Coordinator: Tadeusz Marek


The project regards functioning of neural systems for detecting and correcting perceptual and motor errors. It is aimed at describing patterns of bioelectric brain activity during error processing and correction, as well as at examining the stability of those systems in the chronic sleep deficit state.
The essence of the experiment is simultanous registration of dense array EEG and oculographic data. Various categories of saccadic reactions (correct, erroneous, and corrective ones) will be defined on the basis of precise oculographic recordings. EEG data analysis will be performed according to  those categories.

The former study with functional magnetic resonance (fMRI) technique enabled us to define the localization of brain structures engaged in error detecting and monitoring. The planned research with use of dEEG technique will provide the information about temporal structure of activity within and among brain regions. The laboratory is equipped with a 256-channel EEG system which guarantees the highest possible standard of signal registration, in terms of spatial and temporal resolution (1 milisecond). The system enables a standard analysis (frequency and event-related potentials), as well as localizing of the sources of bioelectric brain activity or implementing other neuroimagining data, e.g. fMRI, and removing artifacts without data lost.

Neural systems for human error processing
EEG and fMRI studies showed that anterior cingulate cortex (ACC) structures are responsible for detecting and correcting errors. Dorsal part of ACC is activated at the cognitive conflict appearance, while activation in the rostral part of ACC emerges after commitment of an error. The effectiveness of these structures may be affected by various individual and situational factors: fatigue, age, alcohol consumption or time characteristics of the task. Sleep deprivation is recognized as one of the most important negative factors which may lead to the total collapse of the system of detection and correction of erroneous responses.  

EEG studies of the event-related potentials (ERP) are focused on analyzing its' various components:
- „error negativity" (Ne) or „error-related negativity" (ERN) – negative polarisation occuring at the moment of error commission, with a peak about 50-100 ms later, considered to be a factor indicating the existence of a discrepancy between the expected response and the realized response;
- "feedback ERN" – the potential recorded in the medial frontal part of the brain, 250-350 ms after an incorrect response and only if a subject is aware of it;
- "error positivity" (Pe) – a component that appears 200–400 ms after an incorrect response, a positive peak registered from the central-parietal part and appearing only in case of error awareness.

Chronic sleep deficit
Sleep curtailment is seen as a hallmark of modern society. In XX century the average sleep has shorten from about 9 hours to 6-7 hours. According to the reports of National Sleep Foundation, Americans sleep on average 44 minutes less than their declared optimal sleep length and 29% complain of chronic sleep deficit, having on average 5 hours 38 minutes on weekdays and 6 hours 26 minutes on weekends. Moreover, during eight years (from 2001 to 2009), the proportion of respondents sleeping less than 6 hours increased from 13% to 20% (on weeknights) and from 7% to 14% (on weekends). At the same time, percentage of people sleeping more than 8 hours decreased, from 38% to 28% and from 61% to 44%, respectively.
Most of laboratory experiments on sleep loss deal with total sleep deprivation. With no doubt, in everyday life, chronic sleep restriction is a much more common occurrence. However, the neurocognitive consequences of chronic partial sleep loss are comparable to those of total sleep deprivation. The effects of sleep restriction comprise uncontrolled microsleeps, unstable attention performance, cognitive slowing which increases errors while time pressure, slowed response time, reduced learning capabilities, deteriorated performance requiring divergent thinking, perseveration of ineffective solutions, general decrease of alertness.
Sleep deprivation augments the risk of human-error accidents. The statistics are alarming: every year in USA 25 000 are killed and 250 000 injured in results of sleep-related accidents. The economic costs are high, estimated as 56 billion dollars yearly. Sleep deficit was recognized as one of important indirect causes of Challenger shuttle catastrophe in 1986, the Chernobyl nuclear plant breakdown, and the oil leak from the Exxon Valdez tanker in 1989.

Disturbances of natural sleep-wake cycle of workers lead to serious risks, especially in case of critical safety operations (nuclear power plants, air traffic control, energetic and chemical industries, public transportation), and also other jobs in which effective error processing is a key issue for human safety (medical services, police, firefighters). Drowsiness at work seems to be a long-known and often ignored hazard. Not before the media alarm the public of dramatic events (e.g. recent reports that several planes couldn't contact airport towers for assistance in landing because of controllers' falling asleep), do those problems get into a focus of interest. Breakdowns in critical safety operations involve huge costs in terms of human, economic, and environmental consequences.

Potential effects and applications of study results
The research on ERP components related to human error monitoring make the potential for constructing BCI (Brain-Computer Interface) systems of a wide range of applications: supporting the training, assistance in the process of rehabilitation, introducing a correction for inadequate behavior and, finally, building early warning systems for the operators of critical safety systems.
Previous research on error processing in a sleep deficit state did not take into account the diurnal variation. They were based on laboratory-controlled total sleep deprivation (many hours in one day) and single measurements. In the planned project the subjects will be induced to chronic (weekly) sleep deficit, in natural conditions, and the efficiency of the functioning (detection and correction of errors) will be monitored four times during the day.
Such a project on the mechanisms of neural processing of errors is part of the latest trend of research focused on accidents prevention. Taking into account the diurnal variability and chronic sleep deprivation in this type of experiment is unique. The findings may have particular importance for working in the area of highly developed technologies.