Omission responses in field potentials but not spikes in rat auditory cortex

Abstract

Non-invasive recordings of gross neural activity in humans often show responses to omitted stimuli in steady trains of identical stimuli. This has been taken as evidence for the neural coding of prediction or prediction error. However, evidence for such omission responses from invasive recordings of cellular-scale responses in animal models is scarce. Here, we sought to characterise omission responses using extracellular recordings in the auditory cortex of anaesthetised rats. We profiled omission responses across local field potentials (LFP), analogue multiunit activity (AMUA), and single/multi-unit spiking activity, using stimuli that were fixed-rate trains of acoustic noise bursts where 5% of bursts were randomly omitted. Significant omission responses were observed in LFP and AMUA signals, but not in spiking activity. These omission responses had a lower amplitude and longer latency than burst-evoked sensory responses, and omission response amplitude increased as a function of the number of preceding bursts. Contrary to theories of neural entrainment, rhythmic stimulus presentation did not increase low-frequency phase-locking of neural activity specific to the stimulus presentation rate. Together, our findings show that omission responses are observed in LFP and AMUA signals, with laminar specificity, but are not observed in spiking activity, and do not show evidence for low-frequency phase locking. This has implications for models of cortical processing that require many neurons to encode prediction error in their spike output, and is consistent with models involving representation of prediction error in apical dendrites electrotonically distant from the soma.

Hierarchical Deviant Processing in Auditory Cortex of Awake Mice

“…Our null findings suggest a number of non-exclusive possibilities. First, it could be that omission responses are not robustly encoded in the widefield calcium activity (analyzed here) or spiking activity mainly from excitatory neurons in the superficial layers (to which wide-field imaging is most sensitive) (Waters, 2020), but rather in population activity (Auksztulewicz et al, 2022) potentially reflecting dendritic currents (Guerguiev et al, 2017), or in other types of neurons. This possibility is consistent with a recent calcium imaging study in the visual cortex, where neural activity following omissions was observed in L2/3 inhibitory neurons, but not in L2/3 excitatory neurons (Garrett et al, 2020).…”

Monkeys have rhythm

“…These include an auditory component capable of abstracting salient temporal patterns out of a complex acoustic signal [33,[51][52][53][54][55][56], a predictive component capable of projecting a pattern forward to predict the timing of future stimulus input [57][58][59][60][61][62], an auditory-motor connection capable of using sensory information and feedback to time motor actions in anticipation of future sound events [63][64][65][66], and the coordination of these through and association with reward that is either intrinsic or conditioned [37][38][39].…”

Predictive Coding Networks for Temporal Prediction

“…Such an architecture can reconcile predictive coding networks with the lack of strong evidence for there being explicit prediction error neurons in the cortex [11], unlike the dopaminergic reward prediction error neurons in the mid-brain whose existence has been established for decades [99, 100]. For example, a recent study in rat cortex found little evidence for prediction-related signals in spikes, but found strong evidence for it in local field potentials, which are thought to be driven by somatic and dendritic potentials [101]. Although in Fig 2B and C we still use error neurons in the hierarchical part of the network, they can also be circumvented following the dendritic implementation in [102].…”