Altered directed connectivity during processing of predictive stimuli in psychiatric patient populations.

OBJECTIVES: The study investigated the role of top-down versus bottom-up connectivity, during the processing of predictive information, in three different psychiatric disorders. METHODS: Electroencephalography (EEG) was recorded during the performance of a task, which evaluates the ability to use predictive information in order to facilitate predictable versus random target detection. We evaluated EEG event-related directed connectivity, in patients with schizophrenia (SZ), major depressive disorder (MDD), and autism spectrum disorder (ASD), compared with healthy age-matched controls. Directed connectivity was evaluated using phase transfer entropy. RESULTS: We showed that top-down frontal-parietal connectivity was weaker in SZ (theta and beta bands) and ASD (alpha band) compared to control subjects, during the processing of stimuli consisting of the predictive sequence. In SZ patients, top-down connectivity was also attenuated, during the processing of predictive targets in the beta frequency band. In contrast, compared with controls, MDD patients displayed an increased top-down flow of information, during the processing of predicted targets (alpha band). CONCLUSIONS: The findings suggest that top-down frontal-parietal connectivity is altered differentially across three major psychiatric disorders, specifically during the processing of predictive stimuli. SIGNIFICANCE: Altered top-down connectivity may contribute to the specific prediction deficits observed in each of the patient populations.

Altered directed connectivity during processing of implicit versus explicit predictive stimuli in Parkinson's disease patients.

The study investigated the role of top-down versus bottom-up connectivity, during the processing of implicit or explicit predictive information, in Parkinson's disease (PD). EEG was recorded during the performance of a task, which evaluated the ability to utilize either implicit or explicit predictive contextual information in order to facilitate the detection of predictable versus random targets. Thus, subjects performed an implicit and explicit session, where subjects were either unaware or made aware of a predictive sequence that signals the presentation of a subsequent target, respectively. We evaluated EEG event-related directed connectivity, in PD patients compared with healthy age-matched controls, using phase transfer entropy. PD patients showed increased top-down frontal-parietal connectivity, compared to control subjects, during the processing of the last (most informative) stimulus of the predictive sequence and of random standards, in the implicit and explicit session, respectively. These findings suggest that PD is associated with compensatory top-down connectivity, specifically during the processing of implicit predictive stimuli. During the explicit session, PD patients seem to allocate more attentional resources to non-informative standard stimuli, compared to controls. These connectivity changes shed further light on the cognitive deficits, associated with the processing of predictive contextual information, that are observed in PD patients.

Functional connectivity abnormalities during processing of predictive stimuli in patients with major depressive disorder.

The study investigated the underlying mechanisms associated with the ability of patients with major depressive disorder (MDD) to utilize predictive contextual information in order to facilitate detection of predictable versus random targets. To this end we evaluated EEG event-related functional connectivity during the processing of predictive stimuli in MDD and control subjects. A target detection task was used where targets were either preceded by randomized sequences of standards, or by sequences that included a predictive sequence. Functional connectivity was evaluated using synchronization likelihood and graph theory. The cluster coefficient and local efficiency values were greater in MDD compared to controls, during the processing of the three stimuli consisting of the predictive sequence, in the beta frequency band, suggesting an increased structured network organization. These changes were associated with increased functional connectivity within frontal networks in MDD patients compared to controls. However, no significant functional connectivity group-changes were observed for target conditions or randomized standards. These findings suggest that MDD is associated with context-specific functional connectivity abnormalities during the processing of predictive stimuli.

Altered predictive contextual processing of emotional faces versus abstract stimuli in adults with Autism Spectrum Disorder.

OBJECTIVES: We investigated the proposition that Autism Spectrum Disorder (ASD) is associated with predictive contextual processing deficits. METHODS: We recorded electroencephalography (EEG) in adults with ASD and controls during the performance of a predictive contextual processing task, using either triangles or emotional faces. Targets were preceded by either randomized sequences (R) or by sequences including a predictive sequence (P). RESULTS: ASD subjects showed an attenuated behavioral facilitation (P versus R) compared with controls (faces). P3b amplitudes of P, R and the predictive sequence (n-1) were attenuated in ASD compared with controls. However, the attenuation of n-1 was more pronounced during the processing of faces. Controls demonstrated shorter peak P3b latencies of P versus R, while this facilitation was absent in ASD subjects. ASD subjects demonstrated functional connectivity alterations during the processing of random (triangles and faces) and predicted targets (faces). These changes were associated with weaker, more randomised, functional connections between frontal and parietal regions in ASD. CONCLUSIONS: We found predictive contextual processing alterations in ASD, which were more pronounced during the processing of emotional faces compared with abstract stimuli. SIGNIFICANCE: We provide novel evidence for the proposition that ASD is associated with deficits of top-down predictions.

Athletes versus video game players: A predictive contextual processing study.

We investigated the effect of abstract and real life meaningful images from sports on predictive contextual processing in professional athletes and video gamers. EEG was recorded in three groups: professional basketball players (BP), professional athletes of individual sports (IA) and experienced action video game players (VG). Two recording sessions, each with a different set of visual stimuli was presented: either triangles facing left, up, right or down or four images of a basketball player throwing a ball. Recording blocks consisted of targets preceded by randomized sequences of standards and by sequences including a predictive sequence signaling the occurrence of a subsequent target event. The gradual increase of P3b amplitudes across the predictive sequence was greater in BP compared with VG, when stimuli consisted of real life images of a basketball player. For the basketball session, we observed increased local modularity and stronger functional connectivity within frontal attentional networks in BP and VG compared with IA, during the processing of the predictive sequence. Our findings suggest increased top-down attentional allocation, during the processing of predictive visual stimuli, in basketball players compared with video gamers and individual sports athletes.

Processing of implicit versus explicit predictive contextual information in Parkinson's disease.

We investigated the effects of implicit versus explicit local contextual processing in Parkinson's disease (PD) using electrophysiological measures. EEG recording blocks consisted of targets preceded by either randomized sequences of standards or by sequences including a predictive sequence signaling the occurrence of a target event. PD patients ("on" medication) and healthy controls performed two sessions: in the first the regularity of the predictive sequence was implicit, while in the second this regularity was made explicit. PD and control subjects showed shorter reaction times for predicted versus random targets across both the implicit and explicit sessions. Healthy controls showed a facilitation of the P3b latency for predicted targets (compared with random targets) in both the implicit and explicit sessions, while PD patients showed no significant P3b latency differences between the two target conditions across both sessions. In healthy controls the P3b amplitude of the last most-informative stimulus of the predictive sequence (n-1) was larger in the explicit compared to the implicit session, while this difference was absent in PD patients. Functional connectivity measures showed that during the explicit and implicit sessions detection of n-1 was associated with higher gamma values and weaker fronto-central-parietal connections, in the theta band, in PD compared to controls. The findings suggest that implicit and explicit processing of predictive contextual information is altered in PD patients and that this may be associated with functional connectivity abnormalities within top-down frontal networks.