188宝金博页面版

Converging intracortical signatures of two separated processing timescalesin human early auditory cortexFabiano Baronia , b , * , 1 , Benjamin Morillon c , Agn? es Tr? ebuchon c , d , Catherine Li? egeois-Chauvel c , e ,Itsaso Olasagastia , Anne-Lise Giraud aa Department of Fundamental Neuroscience, University of Geneva, Geneva, Switzerlandb School of Engineering, ?Ecole Polytechnique F? ed? erale de Lausanne, Lausanne, Switzerlandc Aix Marseille Universit? e, Institut National de la Sant? e et de la Recherche M? edicale (INSERM), Institut de Neurosciences des Syst? emes (INS), Marseille, Franced Clinical Neurophysiology and Epileptology Department, Timone Hospital, Assistance Publique Hôpitaux de Marseille, Marseille, Francee Department of Neurological Surgery, University of Pittsburgh, PA, 15213, USAA R T I C L E I N F OKeywords:iEEGSpeech perceptionSpectral analysisBrain decodingComputational modelingAuditory cortexA B S T R A C TNeural oscillations in auditory cortex are argued to support parsing and representing speech constituents at theircorresponding temporal scales. Yet, how incoming sensory information interacts with ongoing spontaneous brainactivity, what features of the neuronal microcircuitry underlie spontaneous and stimulus-evoked spectral f i n-gerprints, and what these f i ngerprints entail for stimulus encoding, remain largely open questions. We used acombination of human invasive electrophysiology, computational modeling and decoding techniques to assess theinformation encoding properties of brain activity and to relate them to a plausible underlying neuronal micro-architecture. We analyzed intracortical auditory EEG activity from 10 patients while they were listening to shortsentences. Pre-stimulus neural activity in early auditory cortical regions often exhibited power spectra with ashoulder in the delta range and a small bump in the beta range. Speech decreased power in the beta range, andincreased power in the delta-theta and gamma ranges. Using multivariate machine learning techniques, weassessed the spectral prof i le of information content for two aspects of speech processing: detection anddiscrimination. We obtained better phase than power information decoding, and a bimodal spectral prof i le ofinformation content with better decoding at low (delta-theta) and high (gamma) frequencies than at intermediate(beta) frequencies. These experimental data were reproduced by a simple rate model made of two subnetworkswith different timescales, each composed of coupled excitatory and inhibitory units, and connected via a negativefeedback loop. Modeling and experimental results were similar in terms of pre-stimulus spectral prof i le (except forthe iEEG beta bump), spectral modulations with speech, and spectral prof i le of information content. Altogether,we provide converging evidence from both univariate spectral analysis and decoding approaches for a dualtimescale processing infrastructure in human auditory cortex, and show that it is consistent with the dynamics of asimple rate model.1. IntroductionThe brains of humans and other animals generate electrical activitythat often exhibits rhythmic patterns, which are apparent as shoulders orsmall bumps in the power spectrum of electrophysiological signals on topof the 1/f α prof i le (Buzsaki, 2006; Buzs? aki and Draguhn, 2004). It hasbeen suggested that rhythmic activity constitutes the neural basis ofrhythmic and pseudo-rhythmic motor actions such as breathing, loco-motion, chewing, peristalsis and the generation of vocalizations andother communication signals, the latter being more prominently devel-oped in primates and birds.Neural oscillations correspond to rhythmic activity contributed byseveral thousand neurons, astrocytes and possibly other cell types. Theyconstitute a compact, low-dimensional signature of the local networkstate that can be informative about contextual cognitive state (e.g.arousal (Steriade et al., 1993; McGinley et al., 2015), attention (Frieset al., 2001; Besle et al., 2011; Ding and Simon, 2012; Zion Golumbicet al., 2013; Klimesch, 2012; Clayton et al., 2015; Calderone et al., 2014)* Corresponding author. Department of Fundamental Neuroscience, University of Geneva, Geneva, Switzerland.E-mail address: fabianobaroni@gmail.com (F. Baroni).1Current address: Escuela Polit? ecnica Superior, Universidad Aut? onoma de Madrid, Madrid, Spain.Contents lists available at ScienceDirectNeuroImagejournal homepage: www.elsevier.com/locate/neuroimagehttps://doi.org/10.1016/j.neuroimage.2020.116882Received 17 September 2019; Received in revised form 30 March 2020; Accepted 23 April 2020Available online 18 May 20201053-8119/© 2020 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).NeuroImage 218 (2020) 116882