More evidence for a long-latency mismatch response in urethane-anaesthetised mice

Decoding violated sensory expectations from the auditory cortex of anaesthetized mice: Hierarchical recurrent neural network depicts separate ‘danger’ and ‘safety’ units

“…The relationships between tone frequency and obligatory stimulus-on and stimulus-off response peak amplitudes were not proportional, therefore not in agreement with previous findings (Nakamura et al, 2011;O'Reilly and Conway, 2021). Furthermore, long-latency responses were only observed from deviant frequency stimuli exposed to during training, defying the expectation that other frequencies that deviate from the standard would also produce comparable responses in-vivo (Casado-Román et al, 2020;Chen et al, 2015;O'Reilly and Angsuwatanakul, 2021). This suggests that the model has not learned to trigger this response to frequencies that differ from the standard, rather it has learned to produce long-latency responses only to the specific deviant frequency stimuli used during training (7.5 and 12.5 kHz).…”

“…Comparable long-latency responses to frequency oddball stimuli have been observed in anaesthetised rodents (Casado-Román et al, 2020;Chen et al, 2015;O'Reilly and Angsuwatanakul, 2021;Ruusuvirta et al, 1998). Curiously though, these have been absent from studies employing similar or near-identical stimulation and recording protocols in conscious mice (Harms et al, 2014;O'Reilly and Conway, 2021).…”

“…Trials containing two consecutive stimulus responses were extracted; i.e. either responses to two repeated standard stimuli, or responses to an unexpected deviant followed by a standard stimulus, as described previously (O'Reilly, 2019a;O'Reilly and Angsuwatanakul, 2021).…”

“…Ambiguity thereby arises concerning the latency of obligatory and context-dependent components of the auditory response in animals, which may be problematic to dissociate (Parras et al, 2017;Taaseh et al, 2011). Some data from anaesthetised mice suggest that earlier components are influenced more by adaptation and the physical properties of stimuli, whereas later components are context-dependent (Chen et al, 2015;O'Reilly, 2019a;O'Reilly and Angsuwatanakul, 2021). It has also been shown that this late component in anaesthetised mice relies on intact NMDA receptor function (Chen et al, 2015), comparable with human MMN (Avissar and Javitt, 2018;Rosburg and Kreitschmann-Andermahr, 2016).…”

Decoding violated sensory expectations from the auditory cortex of anaesthetised mice: Hierarchical recurrent neural network depicts separate ‘danger’ and ‘safety’ units

“…If these positive amplitude long-latency responses were simply due to the physical makeup of the stimuli, then standard stimulus responses should consistently lie between those of the two deviants, as they do for stimulus onset and offset peaks, but this is not the case. Results from other studies in urethaneanaesthetised mice have also demonstrated that lowprobability stimuli in the oddball sequence, which violate sensory expectations, produce comparable long-latency features that were absent from responses to lowprobability stimuli in the many-standards control sequence, which do not violate an established auditory regularity (Casado-Rom an et al, 2020;Kurkela et al, 2018;O'Reilly & Angsuwatanakul, 2021).…”