Local Deviance Research Articles

Chronic pain – A maladaptive compensation to unbalanced hierarchical predictive processing

The ability to perceive pain presents an interesting evolutionary advantage to adapt to an ever-changing environment. However, in the case of chronic pain (CP), pain perception hinders the capacity of the system to adapt to changing sensory environments. Similar to other chronic perceptual disorders, CP is also proposed to be a maladaptive compensation to aberrant sensory predictive processing. The local-global oddball paradigm relies on learning hierarchical rules and processing environmental irregularities at a local and global level. Prediction errors (PE) between actual and predicted input typically trigger an update of the forward model to limit the probability of encountering future PEs. It has been hypothesised that CP hinders forward model updating, reflected in increased local deviance and decreased global deviance. In the present study, we used the local-global paradigm to examine how CP influences hierarchical learning relative to healthy controls. As hypothesised, we observed that deviance in the stimulus characteristics evoked heightened local deviance and decreased global deviance of the stimulus-driven PE. This is also accompanied by respective changes in theta phase locking that is correlated with the subjective pain perception. Changes in the global deviant in the stimulus-driven-PE could also be explained by dampened attention-related responses. Changing the context of the auditory stimulus did not however show a difference in the context-driven PE. These findings suggest that CP is accompanied by maladaptive forward model updating where the constant presence of pain perception disrupts local deviance in non-nociceptive domains. Furthermore, we hypothesise that the auditory-processing based biomarker identified here could be a marker of domain-general dysfunction that could be confirmed by future research.

Local and Overall Deviance R-Squared Measures for Mixtures of Generalized Linear Models

In generalized linear models (GLMs), measures of lack of fit are typically defined as the deviance between two nested models, and a deviance-based R2 is commonly used to evaluate the fit. In this paper, we extend deviance measures to mixtures of GLMs, whose parameters are estimated by maximum likelihood (ML) via the EM algorithm. Such measures are defined both locally, i.e., at cluster-level, and globally, i.e., with reference to the whole sample. At the cluster-level, we propose a normalized two-term decomposition of the local deviance into explained, and unexplained local deviances. At the sample-level, we introduce an additive normalized decomposition of the total deviance into three terms, where each evaluates a different aspect of the fitted model: (1) the cluster separation on the dependent variable, (2) the proportion of the total deviance explained by the fitted model, and (3) the proportion of the total deviance which remains unexplained. We use both local and global decompositions to define, respectively, local and overall deviance R2 measures for mixtures of GLMs, which we illustrate—for Gaussian, Poisson and binomial responses—by means of a simulation study. The proposed fit measures are then used to assess, and interpret clusters of COVID-19 spread in Italy in two time points.

The Impact of Local Deviance from National-Level Norms on a Community’s Entrepreneurial Activity

Institutional theorists have begun to explore how deviance, or the extent to which institutions are violated, impacts entrepreneurial activity. However, the literature has thus far focused on formal deviance, or the extent to which regulatory institutions are violated, concluding a negative relationship between formal deviance and entrepreneurship due to heightened mistrust in the community. Drawing from Durkheimian Sociology, we postulate that the relationship functions differently for informal deviance, or the extent to which normative institutions are violated. Specifically, we argue that increasing levels of local deviance from national-level norms will foster greater entrepreneurial activity within communities due to informal deviance’s impact on community social capital. We test this argument using longitudinal data on all 2448 municipalities in Mexico. The results, and a series of robustness checks, confirm our hypotheses. Overall, our study contributes to research within the institutional approach to entrepreneurship by demonstrating that the impact of deviance on entrepreneurial activity operates differently across institutional pillars.

Arousal State-Dependence of Interactions Between Short- and Long-Term Auditory Novelty Responses in Human Subjects.

In everyday life, predictable sensory stimuli are generally not ecologically informative. By contrast, novel or unexpected stimuli signal ecologically salient changes in the environment. This idea forms the basis of the predictive coding hypothesis: efficient sensory encoding minimizes neural activity associated with predictable backgrounds and emphasizes detection of changes in the environment. In real life, the brain must resolve multiple unexpected sensory events occurring over different time scales. The local/global deviant experimental paradigm examines auditory predictive coding over multiple time scales. For short-term novelty [hundreds of milliseconds; local deviance (LD)], sequences of identical sounds (/xxxxx/) are interspersed with sequences that contain deviants (/xxxxy/). Long-term novelty [several seconds; global deviance (GD)] is created using either (a) frequent /xxxxx/ and infrequent /xxxxy/ sequences, or (b) frequent /xxxxy/ and infrequent /xxxxx/ sequences. In scenario (a), there is both an LD and a GD effect (LDGD, “double surprise”). In (b), the global deviant is a local standard, i.e., sequence of identical sounds (LSGD). Cortical responses reflecting LD and GD originate in different brain areas, have a different time course, and are differentially sensitive to general anesthesia. Neural processes underlying LD and GD have been shown to interact, reflecting overlapping networks subserving the detection of novel auditory stimuli. This study examined these interactions using intracranial electroencephalography in neurosurgical patients. Subjects performed a GD target detection task before and during induction of anesthesia with propofol. Recordings were made from the auditory cortex, surrounding auditory-related and prefrontal cortex in awake, sedated, and unresponsive states. High gamma activity was used to measure the neural basis of local-by-global novelty interactions. Positive interaction was defined as a greater response to the double surprise LDGD condition compared to LSGD. Negative interaction was defined as a weaker response to LDGD. Positive interaction was more frequent than negative interaction and was primarily found in auditory cortex. Negative interaction typically occurred in prefrontal cortex and was more sensitive to general anesthesia. Temporo-parietal auditory-related areas exhibited both types of interaction. These interactions may have relevance in a clinical setting as biomarkers of conscious perception in the assessment of depth of anesthesia and disorders of consciousness.

Cortical responses to auditory novelty across task conditions: An intracranial electrophysiology study

Elucidating changes in sensory processing across attentional and arousal states is a major focus in neuroscience. The local/global deviant (LGD) stimulus paradigm engages auditory predictive coding over short (local deviance, LD) and long (global deviance, GD) time scales, and has been used to assay disruption of auditory predictive coding upon loss of consciousness. Our previous work (Nourski et al., 2018, J Neurosci 38:8441–52) examined effects of general anesthesia on short- and long-term novelty detection. GD effects were suppressed at subhypnotic doses of propofol, suggesting that they may be more related to task engagement than consciousness per se. The present study addressed this hypothesis by comparing cortical responses to auditory novelty during passive versus active listening conditions in awake listeners. Subjects were seven adult neurosurgical patients undergoing chronic invasive monitoring for medically intractable epilepsy. LGD stimuli were sequences of four identical vowels followed by a fifth identical or different vowel. In the passive condition, the stimuli were presented to subjects as they watched a silent TV program and were instructed to attend to its content. In the active condition, stimuli were presented in the absence of a TV program, and subjects were instructed to press a button in response to GD target stimuli. Intracranial recordings were made from multiple brain regions, including core and non-core auditory, auditory-related, prefrontal and sensorimotor cortex. Metrics of task performance included hit rate, sensitivity index, and reaction times. Cortical activity was measured as averaged auditory evoked potentials (AEPs) and event-related band power in high gamma (70–150 Hz) and alpha (8–14 Hz) frequency bands. The vowel stimuli and LD elicited robust AEPs in all studied brain areas in both passive and active conditions. High gamma responses to stimulus onset and LD were localized predominantly to the auditory cortex in the superior temporal plane and had a comparable prevalence and spatial extent between the two conditions. In contrast, GD effects (AEPs, high gamma and alpha suppression) were greatly enhanced during the active condition in all studied brain areas. The prevalence of high gamma GD effects was positively correlated with individual subjects’ task performance. The data demonstrate distinct task engagement-related effects on responses to auditory novelty across the auditory cortical processing hierarchy. The results motivate a closer examination of effective connectivity underlying attentional modulation of cortical sensory responses, and serve as a foundation for examining changes in sensory processing associated with general anesthesia, sleep and disorders of consciousness.

Auditory Predictive Coding across Awareness States under Anesthesia: An Intracranial Electrophysiology Study.

The systems-level mechanisms underlying loss of consciousness (LOC) under anesthesia remain unclear. General anesthetics suppress sensory responses within higher-order cortex and feedback connections, both critical elements of predictive coding hypotheses of conscious perception. Responses to auditory novelty may offer promise as biomarkers for consciousness. This study examined anesthesia-induced changes in auditory novelty responses over short (local deviant [LD]) and long (global deviant [GD]) time scales, envisioned to engage preattentive and conscious levels of processing, respectively. Electrocorticographic recordings were obtained in human neurosurgical patients (3 male, 3 female) from four hierarchical processing levels: core auditory cortex, non-core auditory cortex, auditory-related, and PFC. Stimuli were vowel patterns incorporating deviants within and across stimuli (LD and GD). Subjects were presented with stimuli while awake, and during sedation (responsive) and following LOC (unresponsive) under propofol anesthesia. LD and GD effects were assayed as the averaged evoked potential and high gamma (70-150 Hz) activity. In the awake state, LD and GD effects were present in all recorded regions, with averaged evoked potential effects more broadly distributed than high gamma activity. Under sedation, LD effects were preserved in all regions, except PFC. LOC was accompanied by loss of LD effects outside of auditory cortex. By contrast, GD effects were markedly suppressed under sedation in all regions and were absent following LOC. Thus, although the presence of GD effects is indicative of being awake, its absence is not indicative of LOC. Loss of LD effects in higher-order cortical areas may constitute an alternative biomarker of LOC.SIGNIFICANCE STATEMENT Development of a biomarker that indexes changes in the brain upon loss of consciousness (LOC) under general anesthesia has broad implications for elucidating the neural basis of awareness and clinical relevance to mechanisms of sleep, coma, and disorders of consciousness. Using intracranial recordings from neurosurgery patients, we investigated changes in the activation of cortical networks involved in auditory novelty detection over short (local deviance) and long (global deviance) time scales associated with sedation and LOC under propofol anesthesia. Our results indicate that, whereas the presence of global deviance effects can index awareness, their loss cannot serve as a biomarker for LOC. The dramatic reduction of local deviance effects in areas beyond auditory cortex may constitute an alternative biomarker of LOC.

Processing of auditory novelty across the cortical hierarchy: An intracranial electrophysiology study

Under the predictive coding hypothesis, specific spatiotemporal patterns of cortical activation are postulated to occur during sensory processing as expectations generate feedback predictions and prediction errors generate feedforward signals. Establishing experimental evidence for this information flow within cortical hierarchy has been difficult, especially in humans, due to spatial and temporal limitations of non-invasive measures of cortical activity. This study investigated cortical responses to auditory novelty using the local/global deviant paradigm, which engages the hierarchical network underlying auditory predictive coding over short (‘local deviance’; LD) and long (‘global deviance’; GD) time scales. Electrocorticographic responses to auditory stimuli were obtained in neurosurgical patients from regions of interest (ROIs) including auditory, auditory-related and prefrontal cortex. LD and GD effects were assayed in averaged evoked potential (AEP) and high gamma (70–150 Hz) signals, the former likely dominated by local synaptic currents and the latter largely reflecting local spiking activity. AEP LD effects were distributed across all ROIs, with greatest percentage of significant sites in core and non-core auditory cortex. High gamma LD effects were localized primarily to auditory cortex in the superior temporal plane and on the lateral surface of the superior temporal gyrus (STG). LD effects exhibited progressively longer latencies in core, non-core, auditory-related and prefrontal cortices, consistent with feedforward signaling. The spatial distribution of AEP GD effects overlapped that of LD effects, but high gamma GD effects were more restricted to non-core areas. High gamma GD effects had shortest latencies in STG and preceded AEP GD effects in most ROIs. This latency profile, along with the paucity of high gamma GD effects in the superior temporal plane, suggest that the STG plays a prominent role in initiating novelty detection signals over long time scales. Thus, the data demonstrate distinct patterns of information flow in human cortex associated with auditory novelty detection over multiple time scales.

Evaluation of cluster recovery for small area relative risk models

The analysis of disease risk is often considered via relative risk. The comparison of relative risk estimation methods with "true risk" scenarios has been considered on various occasions. However, there has been little examination of how well competing methods perform when the focus is clustering of risk. In this paper, a simulated evaluation of a range of potential spatial risk models and a range of measures that can be used for (a) cluster goodness of fit, (b) cluster diagnostics are considered. Results suggest that exceedence probability is a poor measure of hot spot clustering because of model dependence, whereas residual-based methods are less model dependent and perform better. Local deviance information criteria measures perform well, but conditional predictive ordinate measures yield a high false positive rate.

The diffusion of racist violence in the Netherlands: Discourse and distance

This article illuminates the unanticipated but intense waves of xenophobia that have swept through Western Europe over the last decade. The author makes use of a unique dataset and diffusion models to simultaneously investigate the geographical and temporal development of waves of racist violence in the Netherlands during the turbulent period 2001–03, when the country lost its reputation as a multicultural paradise. The results provide evidence for the fact that previous riots enhance the legitimacy of violence elsewhere, especially if they are visible in the mass media, resonate with public debates on immigration and take place in nearby regions. Opposing previous research on mobilization, the analysis suggests that proxies for ethnic competition, deprivation and political opportunity structures are not significantly related to the outbreak of violence; only population size adequately predicts where violence starts. Together these findings suggest that waves of xenophobia develop in two steps: they start in large cities and subsequently spread to nearby places through geographically clustered networks and to more distant counties once they become visible and resonate in the mass media, turning violence from local deviance into a supra-local phenomenon. This process sheds light on how scales of protest shift and explains why seemingly tolerant regions can suddenly become xenophobic hotbeds.

Some diagnostics for binary logistic regression via smoothing

SUMMARY The paper discusses some diagnostic tools for binary logistic regression which use smoothing techniques. Smoothed binary data are employed to devise a battery of diagnostic plots, from simple procedures which indicate only the adequacy or inadequacy of a maximum likelihood logistic fit, to partial residual plots and delta local deviance plots which reveal the type of model inadequacy, suggesting perhaps the addition of nonlinear terms or interactions. Properties of the methods are studied by simulation. The diagnostic tools are illustrated on a set of data in which the probability of having telephone service is related to the demographic variables income and education.

The Fear of Crime in a Small Town

There has been in American history a myth that attributes virtue to small town life and evil and corruption to urban life. This myth has been on the decline in "Delta City,"* a small industrial town in California. The traditional myth allowed Delta City adults to perceive the source of deviance in their community in either external urban influences or in certain limited local groups ("hoods'). Their sudden realization that there was widespread drug use among youth constituted the first vital blow to this myth. This awareness opened a "Pandora's box," producing considerable confusion and anxiety over locat ing the sources of local deviance. The community's confusion and fear were reflected in its response to youthful drug use and crime.