Validation of non-REM sleep stage decoding from resting state fMRI using linear support vector machines.

A growing body of literature suggests that changes in consciousness are reflected in specific connectivity patterns of the brain as obtained from resting state fMRI (rs-fMRI). As simultaneous electroencephalography (EEG) is often unavailable, decoding of potentially confounding sleep patterns from rs-fMRI itself might be useful and improve data interpretation. Linear support vector machine classifiers were trained on combined rs-fMRI/EEG recordings from 25 subjects to separate wakefulness (S0) from non-rapid eye movement (NREM) sleep stages 1 (S1), 2 (S2), slow wave sleep (SW) and all three sleep stages combined (SX). Classifier performance was quantified by a leave-one-subject-out cross-validation (LOSO-CV) and on an independent validation dataset comprising 19 subjects. Results demonstrated excellent performance with areas under the receiver operating characteristics curve (AUCs) close to 1.0 for the discrimination of sleep from wakefulness (S0|SX), S0|S1, S0|S2 and S0|SW, and good to excellent performance for the classification between sleep stages (S1|S2:~0.9; S1|SW:~1.0; S2|SW:~0.8). Application windows of fMRI data from about 70 s were found as minimum to provide reliable classifications. Discrimination patterns pointed to subcortical-cortical connectivity and within-occipital lobe reorganization of connectivity as strongest carriers of discriminative information. In conclusion, we report that functional connectivity analysis allows valid classification of NREM sleep stages.

The neural correlates of negative prediction error signaling in human fear conditioning.

In a temporal difference (TD) learning approach to classical conditioning, a prediction error (PE) signal shifts from outcome deliverance to the onset of the conditioned stimulus. Omission of an expected outcome results in a negative PE signal, which is the initial step towards successful extinction. In order to visualize negative PE signaling during fear conditioning, we employed combined functional magnetic resonance (fMRI) and skin conductance response (SCR) measurements in a conditioning task with visual stimuli and mild electrical shocks. Positive PE signaling was associated with increased activation in the bilateral insula, supplementary motor area, brainstem, and visual cortices. Negative PE signaling was associated with increased activation in the ventromedial and dorsolateral prefrontal cortices, the left lateral orbital gyrus, the middle temporal gyri, angular gyri, and visual cortices. The involvement of the ventromedial prefrontal and orbitofrontal cortex in extinction learning has been well documented, and this study provides evidence for the notion that these regions are already involved in negative PE signaling during fear conditioning.

The neural correlates and temporal sequence of the relationship between shock exposure, disturbed sleep and impaired consolidation of fear extinction.

Consolidation of extinction learning is a primary mechanism disrupted in posttraumatic stress disorder (PTSD), associated with hypoactivity of the ventromedial prefrontal cortex and hippocampus. A role for rapid eye movement (REM) sleep disturbances in this failure to consolidate extinction learning has been proposed. We performed functional magnetic resonance imaging (fMRI) with simultaneous skin conductance response (SCR) measurements in 16 healthy participants during conditioning/extinction and later recall of extinction. The visual stimuli were basic geometric forms and electrical shocks functioned as the unconditioned stimulus. Between the conditioning/extinction and recall sessions, participants received a 90-min sleep window in the sleep laboratory. This daytime sleep was polysomnographically recorded and scored by professionals blind to the study design. Only seven out of 16 participants had REM sleep; participants without REM sleep had a significantly slower decline of both SCR and neural activity of the laterodorsal tegmentum in response to electrical shocks during conditioning. At recall of fear extinction, participants with preceding REM sleep had a reduced SCR and stronger activation of the left ventromedial prefrontal cortex and bilateral lingual gyrus in response to the extinguished stimulus than participants lacking REM sleep. This study indicates that trait-like differences in shock reactivity/habituation (mediated by the brainstem) are predictive of REM sleep disruption, which in turn is associated with impaired consolidation of extinction (mediated by the ventromedial prefrontal cortex). These findings help understand the neurobiological basis and the temporal sequence of the relationship between shock exposure, disturbed sleep and impaired consolidation of extinction, as observed in PTSD.