Cortical ensembles selective for context

Abstract

Neural processing of sensory information is strongly influenced by context. For instance, cortical responses are reduced to predictable stimuli, while responses are increased to novel stimuli that deviate from contextual regularities. Such bidirectional modulation based on preceding sensory context is likely a critical component or manifestation of attention, learning, and behavior, yet how it arises in cortical circuits remains unclear. Using volumetric two-photon calcium imaging and local field potentials in primary visual cortex (V1) from awake mice presented with visual “oddball” paradigms, we identify both reductions and augmentations of stimulus-evoked responses depending, on whether the stimulus was redundant or deviant, respectively. Interestingly, deviance-augmented responses were limited to a specific subset of neurons mostly in supragranular layers. These deviance-detecting cells were spatially intermixed with other visually responsive neurons and were functionally correlated, forming a neuronal ensemble. Optogenetic suppression of prefrontal inputs to V1 reduced the contextual selectivity of deviance-detecting ensembles, demonstrating a causal role for top-down inputs. The presence of specialized context-selective ensembles in primary sensory cortex, modulated by higher cortical areas, provides a circuit substrate for the brain’s construction and selection of prediction errors, computations which are key for survival and deficient in many psychiatric disorders.

Corticothalamic Pathways in Auditory Processing: Recent Advances and Insights From Other Sensory Systems

“…If SSA neurons are indeed error units, they should be accompanied by activity of neurons encoding expectancy in upper hierarchical levels (see Fiser et al, 2016 , for an elegant protocol and experiment of this kind). We do not yet have such evidence for SSA (but see Hamm et al, 2021 ). It remains an open question if the effects observed using the oddball paradigm in single units are a consequence of perceptual expectations.…”

“…Responses to the deviant are maximal in prefrontal cortex (Casado-Román et al, 2020 ). Because their enhanced responses to surprising sounds cannot be explained solely by bottom-up mechanisms of neural fatigue, these neurons are believed to signal expectancy deviance that can be regarded as a prediction error signal (Parras et al, 2017 ; Malmierca et al, 2019 ; Hamm et al, 2021 ). This enhancement would be the result of modulatory mechanisms exerted by higher order cortical areas (e.g., prefrontal cortex; Hamm et al, 2021 ) and/or neuromodulatory gain mechanisms compatible with precision-weighting (e.g., dopaminergic neuromodulation; Valdés-Baizabal et al, 2020 ).…”

“…However, some of these prediction error units (high SSA) in the thalamus are facilitated by the cortex ( Figure 2B ; Antunes and Malmierca, 2011 ). Hypothetically, this facilitation could be the result of precision-weighting gain mechanisms imposed by CT pathways, probably elicited by the unexpected appearance of a deviant sound (Hamm et al, 2021 ). However, in our opinion, the discussion of whether SSA neurons are prediction error units in terms of predictive coding is highly speculative.…”

Spontaneously emerging internal models of visual sequences combine abstract and event-specific information in the prefrontal cortex

“…In the mouse primary visual cortex, neurons perform deviance detection in correlation with feature (i.e. orientation) selectivity (Hamm et al, 2021). Therefore, different V1 neurons respond to deviant stimuli depending on their tuning.…”

Long- and short-term history effects in a spiking network model of statistical learning

“…Firstly, a priori knowledge of sensory stimuli influences perception. For example, an expected stimulus might be encoded faster than an unexpected stimulus (Kok et al 2017; Mazzucato et al 2019), conversely, low-probability stimuli might trigger a stronger response to signal novelty or surprise (Ulanovsky et al 2003; Khouri and Nelken 2015; Hamm et al 2021). Secondly, knowledge of the statistics of a relevant variable influences decisions, potentially leading to biases (Akrami et al 2018; Zylberberg et al 2018; Lieder et al 2019; Hachen et al 2020; Meirhaeghe et al 2021).…”