Biology subjects

Miller, E. K.

Publications and source records attributed to Miller, E. K..

At least 37 records · Page 2Linked to original sources

bioRxiv2024

Convergent effects of different anesthetics are due to changes in phase alignment of cortical oscillations

Many anesthetics cause loss of responsiveness despite having diverse underlying molecular and circuit actions. To explore the convergent effects of these drugs, we examined how anesthetic doses of ketamine and dexmedetomidine affected oscillations in the prefrontal cortex of nonhuman primates. Both anesthetics caused increases in phase locking in the ventrolateral and dorsolateral prefrontal cortex, within and across hemispheres. However, the nature of the phase locking varied. Activity in different subregions within a hemisphere became more anti-phase with both drugs. Local analyses within a region suggested that this finding could be explained by broad cortical distance-based effects, such as large traveling waves. By contrast, homologous areas across hemispheres became more in-phase. Our results suggest that both anesthetics induce strong patterns of cortical phase alignment that are markedly different from those in the awake state, and that these patterns may be a common feature driving loss of responsiveness from different anesthetic drugs.

neuroscience↗

bioRxiv2024

Stability from subspace rotations and traveling waves

Cortical activity shows the ability to recover from distractions. We analyzed neural activity from the prefrontal cortex (PFC) of monkeys performing working memory tasks with mid-memory-delay distractions (a cued gaze shift or an irrelevant visual input). After distraction there were state-space rotational dynamics that returned spiking to population patterns similar to those pre-disruption. In fact, rotations were fuller when the task was performed correctly versus when errors were made. We found a correspondence between state-space rotations and traveling waves across the surface of the PFC. This suggests a role for emergent dynamics like state-space rotations and traveling waves in recovery from distractions.

neuroscience↗

bioRxiv2023

Adaptive stretching of representations across brain regions and deep learning model layers

Prefrontal cortex (PFC) is known to modulate the visual system to favor goal-relevant information by accentuating task-relevant stimulus dimensions. Does the brain broadly re-configures itself to optimize performance by stretching visual representations along task-relevant dimensions? We considered a task that required monkeys to selectively attend on a trial-by-trial basis to one of two dimensions (color or motion direction) to make a decision. Except for V4 (color bound) and MT (motion bound), the brain radically re-configured itself to stretch representations along task-relevant dimensions in lateral PFC, frontal eye fields (FEF), lateral intraparietal cortex (LIP), and inferotemporal cortex (IT). Spike timing was crucial to this code. A deep learning model was trained on the same visual input and rewards as the monkeys. Despite lacking an explicit selective attention or other control mechanism, the model displayed task-relevant stretching as a consequence of error minimization, indicating that stretching is an adaptive strategy.

neuroscience↗

bioRxiv2023

Propofol anesthesia destabilizes neural dynamics across cortex

Every day, hundreds of thousands of people undergo general anesthesia. One hypothesis is that anesthesia disrupts dynamic stability, the ability of the brain to balance excitability with the need to be stable and thus controllable. We tested this hypothesis using a new method for quantifying population-level dynamic stability in complex systems, Delayed Linear Analysis for Stability Estimation (DeLASE). Propofol was used to transition animals between the awake state and anesthetized unconsciousness. DeLASE was applied to macaque cortex local field potentials (LFPs). We found that neural dynamics were more unstable in unconsciousness compared to the awake state. Cortical trajectories mirrored predictions from destabilized linear systems. We mimicked the effect of propofol in simulated neural networks by increasing inhibitory tone. Paradoxically, increased inhibition also destabilized the networks. Our results suggest that anesthesia disrupts dynamical stability that is required for consciousness.

neuroscience↗

bioRxiv2023

New learning principles emerge from biomimetic computational primitives

Although computational models have deepened our understanding of neuroscience, it is still highly challenging to link actual low-level physiological activity (spiking, field potentials) and biochemistry (transmitters and receptors) with high-level cognitive abilities (decision-making, working memory) nor with corresponding disorders. We introduce an anatomically-organized multi-scale model directly generating simulated physiology from which extended neural and cognitive phenomena emerge. The model produces spiking, fields, phase synchronies, and synaptic change, directly generating working memory, decisions, and categorization, all of which were then validated on extensive experimental macaque data from which the model received zero prior training of any kind. Moreover, the simulation uncovered a previously unknown neural code specifically predicting upcoming erroneous ("incongruous") behaviors, also subsequently confirmed in empirical data. The biomimetic model thus directly and predictively links novel decision and reinforcement signals, of computational interest, with novel spiking and field codes, of potential behavioral and clinical relevance.

neuroscience↗

bioRxiv2023

Propofol-mediated loss of consciousness disrupts predictive routing and local field phase modulation of neural activity

Predictive coding is a fundamental function of the cortex. The predictive routing model proposes a neurophysiological implementation for predictive coding. Predictions are fed back from deep-layer cortex via alpha/beta (8-30Hz) oscillations. They inhibit the gamma (40-100Hz) and spiking that feed sensory inputs forward. Unpredicted inputs arrive in circuits unprepared by alpha/beta, resulting in enhanced gamma and spiking. To test the predictive routing model and its role in consciousness, we collected data from intracranial recordings of macaque monkeys during passive presentation of auditory oddballs (e.g., AAAAB) before and after propofol-mediated loss of consciousness (LOC). In line with the predictive routing model, alpha/beta oscillations in the awake state served to inhibit the processing of predictable stimuli. Propofol-mediated LOC eliminated alpha/beta modulation by a predictable stimulus in sensory cortex and alpha/beta coherence between sensory and frontal areas. As a result, oddball stimuli evoked enhanced gamma power, late (> 200 ms from stimulus onset) period spiking, and superficial layer sinks in sensory cortex. Therefore, auditory cortex was in a disinhibited state during propofol-mediated LOC. However, despite these enhanced feedforward responses in auditory cortex, there was a loss of differential spiking to oddballs in higher order cortex. This may be a consequence of a loss of within-area and inter-area spike-field coupling in the alpha/beta and gamma frequency bands. These results provide strong constraints for current theories of consciousness. Significance statementNeurophysiology studies have found alpha/beta oscillations (8-30Hz), gamma oscillations (40-100Hz), and spiking activity during cognition. Alpha/beta power has an inverse relationship with gamma power/spiking. This inverse relationship suggests that gamma/spiking are under the inhibitory control of alpha/beta. The predictive routing model hypothesizes that alpha/beta oscillations selectively inhibit (and thereby control) cortical activity that is predictable. We tested whether this inhibitory control is a signature of consciousness. We used multi-area neurophysiology recordings in monkeys presented with tone sequences that varied in predictability. We recorded brain activity as the anesthetic propofol was administered to manipulate consciousness. Compared to conscious processing, propofol-mediated unconsciousness disrupted alpha/beta inhibitory control during predictive processing. This led to a disinhibition of gamma/spiking, consistent with the predictive routing model.

neuroscience↗

bioRxiv2023

Propofol Mediated Unconsciousness Disrupts Progression of Sensory Signals through the Cortical Hierarchy

A critical component of anesthesia is the loss sensory perception. Propofol is the most widely used drug for general anesthesia, but the neural mechanisms of how and when it disrupts sensory processing are not fully understood. We analyzed local field potential (LFP) and spiking recorded from Utah arrays in auditory cortex, associative cortex, and cognitive cortex of non-human primates before and during propofol mediated unconsciousness. Sensory stimuli elicited robust and decodable stimulus responses and triggered periods of stimulus-induced coherence between brain areas in the LFP of awake animals. By contrast, propofol mediated unconsciousness eliminated stimulus-induced coherence and drastically weakened stimulus responses and information in all brain areas except for auditory cortex, where responses and information persisted. However, we found stimuli occurring during spiking Up states triggered weaker spiking responses than in awake animals in auditory cortex, and little or no spiking responses in higher order areas. These results suggest that propofols effect on sensory processing is not just due to asynchronous down states. Rather, both Down states and Up states reflect disrupted dynamics.

neuroscience↗

bioRxiv2023

Coding of time with non-linear mixed selectivity in Prefrontal Cortex ensembles

Previous work has identified stimulus specific time cells as a potential mechanism for working memory maintenance. It has been proposed that populations of stimulus specific sequences of cells could support memory for many items in a list over long periods of time. This would require information about one stimulus to persist after the presentation of subsequent stimuli. However, it is not known if sequences triggered by one stimulus persist past the presentation of additional stimuli. It is possible that each new stimulus terminates preceding sequences, making memory for multiple stimuli impossible. To investigate this question, we utilized a data set originally published by (Warden & Miller, 2010), studying the firing of monkey prefrontal neurons during short lists of stimuli. We were able to decode "what happened when" throughout the list, using linear discriminant analysis. Additionally, we were able to decode the first stimulus after the presentation of the second stimulus. Furthermore, we found that stimulus modulated sequences of cells, with discrete temporal fields, continue after the second item was presented. Much of the information about the previous item was carried by neurons that responded to conjunctions of stimuli and the timing of late-firing cells was synchronized to the firing of the second stimulus rather than the first. These properties falsify a simple linear model of sequential time cells. These results suggest that non-linear mixed selectivity extends to continuous variables such as time, but that in this experiment at least, only the timing of the most recent stimulus was explicitly maintained in ongoing firing.

neuroscience↗

bioRxiv2023

In vivo ephaptic coupling allows memory network formation

It is increasingly clear that memories are distributed across multiple brain areas. Such "engram complexes" are important features of memory formation and consolidation. Here, we test the hypothesis that engram complexes are formed in part by bioelectric fields that sculpt and guide the neural activity and tie together the areas that participate in engram complexes. Like the conductor of an orchestra, the fields influence each musician or neuron and orchestrate the output, the symphony. Our results use the theory of synergetics, machine learning and data from a spatial delayed saccade task and provide evidence for in vivo ephaptic coupling in memory representations.

neuroscience↗

bioRxiv2023

Measurable fields to spikes causality and its dependence on cortical layer and area.

Distinct dynamics in different cortical layers are apparent in neuronal and local field potential (LFP) patterns, yet their associations in the context of laminar processing have been sparingly analyzed. Here, we study the laminar organization of spike-field causal flow within and across visual (V4) and frontal areas (PFC) of monkeys performing a visual task. Using an event-based quantification of LFPs and a directed information estimator, we found area and frequency specificity in the laminar organization of spike-field causal connectivity. Gamma bursts (40-80 Hz) in the superficial layers of V4 largely drove intralaminar spiking. These gamma influences also fed forward up the cortical hierarchy to modulate laminar spiking in PFC. In PFC, the direction of intralaminar information flow was from spikes [->] fields where these influences dually controlled top-down and bottom-up processing. Our results, enabled by innovative methodologies, emphasize the complexities of spike-field causal interactions amongst multiple brain areas and behavior.

neuroscience↗

bioRxiv2022

Multifunctional fibers enable modulation of cortical and deep brain activity during cognitive behavior in macaques

Recording and modulating neural activity in vivo enables investigations of neural circuits during behavior. However, there is a dearth of tools for simultaneous recording and localized receptor modulation in large animal models. We address this limitation by translating multifunctional fiber-based neurotechnology previously only available for rodent studies to enable cortical and subcortical neural modulation in macaques. We record single unit and local field potential activity before, during, and after intracranial GABA infusions in the premotor cortex and putamen. We apply state-space models to characterize changes in neural activity and investigate how neural activity evoked by a working memory task varies in the presence of local inhibition. The recordings provide detailed insight into the electrophysiological effect of neurotransmitter receptor modulation in both cortical and subcortical structures in an awake, behaving macaque. Our results demonstrate a first-time translation of multifunctional fibers for causal studies in behaving non-human primates.

neuroscience↗

bioRxiv2022

A ubiquitous spectrolaminar motif of local field potential power across the primate cortex

The mammalian cerebral cortex is anatomically organized into a six-layer motif. It is currently unknown whether a corresponding laminar motif of neuronal activity exists across the cortex. Here, we report such a motif in the power of local field potentials (LFP). We implanted multicontact laminar probes in five macaque monkeys and recorded activity across layers in 14 cortical areas at various hierarchical processing stages in all cortical lobes. The anatomical laminar locations of recordings were histologically identified via electrolytic lesions. In all areas, we found a common spectrolaminar pattern characterized by an increasing deep-to-superficial layer gradient of gamma frequency LFP power peaking in layers 2/3, and an increasing superficial-to-deep gradient of alpha-beta power peaking in layers 5/6. Our results show an electrophysiological dissociation between superficial and deep layers that is preserved across the cortex, suggesting a ubiquitous layer and frequency-based mechanism for cortical computation.

neuroscience↗

bioRxiv2022

A physical neural mass model framework for the analysis of oscillatory generators from laminar electrophysiological recordings

Cortical function emerges from the interactions of multi-scale networks that may be studied at a high level using neural mass models (NMM), representing the mean activity of large numbers of neurons. In order to properly reproduce experimental data, these models require the addition of further elements. Here we provide a framework integrating conduction physics that can be used to simulate cortical electrophysiology measurements, particularly those obtained from multi-contact laminar electrodes. This is achieved by endowing NMMs with basic physical properties, such as the average laminar location of the apical and basal dendrites of pyramidal cell populations. We call this framework laminar NMM, or LaNMM for short. We then employ this framework to infer the location of oscillatory generators from laminar-resolved data collected from the prefrontal cortex in the macaque monkey. Based on the literature on columnar connectivity, we define a minimal neural mass model capable of generating amplitude and phase coupled slow (alpha/beta, 4-22 Hz) and fast (gamma, 30-250 Hz) oscillations. The synapse layer locations of the two pyramidal cell populations are treated as optimization parameters, together with two more LaNMM-specific parameters, to compare the models with the multi-contact recordings. We rank the candidate models using an optimization function that evaluates the match between the functional connectivity (FC) of the model and data, where the FC is defined by the covariance between bipolar voltage measurements at different cortical depths. The family of best solutions reproduces the FC of the observed electrophysiology while selecting locations of pyramidal cells and their synapses that result in the generation of fast activity at superficial layers and slow activity across most depths, in line with recent literature proposals. HighlightsO_LIWe provide a neural mass modeling formalism that includes a physical layer to simulate electrophysiology measurements. C_LIO_LITo analyze in-vivo data collected in the macaque monkey during a memory task, we propose a specific model with two coupled main circuits that can generate realistic electrophysiological signals in two important oscillatory regimes--the alpha/beta and the gamma bands. C_LIO_LIPhysical elements in the model shed light on the generation of oscillations in the two regimes and on the relative power distribution of fast and slow oscillatory signals across cortical depth, which we show can be altered by the choice of the reference location or method. C_LIO_LIThe model is contrasted with in-vivo data, with parameters adjusted by matching voltage statistics in the alpha/beta and gamma bands, leading to a solution with slow frequency components generated by synapses spanning most cortical layers and fast oscillations in superficial layers. C_LIO_LIThe resulting formalism provides useful tools and concepts to analyze and model data, with implications for understanding altered oscillatory EEG activity in dementia, Alzheimers disease and other disorders with oscillatory features. C_LI

neuroscience↗

bioRxiv2022

Reduced variability of bursting activity during working memory

Working memories have long been thought to be maintained by persistent spiking. However, mounting evidence from multiple-electrode recording (and single-trial analyses) shows that the underlying spiking is better characterized by intermittent bursts of activity. A counterargument suggested this intermittent activity is at odds with observations that spike-time variability reduces during task performance. However, this counterargument rests on assumptions, such as randomness in the timing of the bursts, that may not be correct. Thus, we analyzed spiking and LFPs from the prefrontal cortex (PFC) of monkeys to determine if task-related reductions in variability can co-exist with intermittent spiking. We found that it does because both spiking and associated gamma bursts were task-modulated, not random. In fact, the task-related reduction in spike variability could be explained by a related reduction in gamma burst variability. Our results provide further support for the intermittent activity models of working memory as well as novel mechanistic insights into how spike variability is reduced during cognitive tasks.

neuroscience↗

bioRxiv2022

Propofol anesthesia alters cortical traveling waves

Oscillatory dynamics in cortex seem to organize into traveling waves that serve a variety of functions. Recent studies show that propofol, a widely used anesthetic, dramatically alters cortical oscillations by increasing slow-delta oscillatory power and coherence. It is not known how this affects traveling waves. We compared traveling waves across the cortex of non-human primates (NHPs) before, during, and after propofol-induced loss-of-consciousness (LOC). After LOC, traveling waves in the slow-delta ([~] 1Hz) range increased, grew more organized, and travelled in different directions relative to the awake state. Higher frequency (8-30 Hz) traveling waves, by contrast, decreased, lost structure, and switched to directions where the slow-delta waves were less frequent. The results suggest that LOC may be due, in part, to changes in slow-delta traveling waves that, in turn, alter and disrupt traveling waves in the higher frequencies associated with cognition.

neuroscience↗

bioRxiv2022

Robust Working Memory through Short-Term Synaptic Plasticity

Working memory has long been thought to arise from sustained spiking/attractor dynamics. However, recent work has suggested that short-term synaptic plasticity (STSP) may help maintain attractor states over gaps in time with little or no spiking. To determine if STSP endows additional functional advantages, we trained artificial recurrent neural networks (RNNs) with and without STSP to perform an object working memory task. We found that RNNs with and without STSP were able to maintain memories over distractors presented in the middle of the memory delay. However, RNNs with STSP showed activity that was similar to that seen in the cortex of a non-human primate (NHP) performing the same task. By contrast, RNNs without STSP showed activity that was less brain-like. Further, RNNs with STSP were more robust to noise and network degradation than RNNs without STSP. These results show that STSP can not only help maintain working memories, it also makes neural networks more robust and brain-like.

neuroscience↗

bioRxiv2021

Closed-Loop Control of Anesthetic State in Non-Human Primates

Continuous monitoring of electroencephalogram (EEG) recordings in humans under general anesthesia (GA) has demonstrated that changes in EEG dynamics induced by an anesthetic drug are reliably associated with the altered arousal states caused by the drug. This observation suggests that an intelligent, closed-loop anesthesia delivery (CLAD) system operating in real-time could track EEG dynamics and control the infusion rate of a programmable pump to precisely maintain unconsciousness. The United States FDA acknowledges the potential benefits of such automatic physiological closed-loop control devices for patient care. Bringing these devices into clinical practice requires establishing their feasibility in suitable animal models. Therefore, given the close neurophysiological proximity between human and non-human primates (NHPs), we address this problem by developing and validating a propofol CLAD system in rhesus macaques. Our CLAD system has three key components: (1) a data acquisition system that records cortical local field potentials (LFPs) from an NHP in real-time; (2) a computer executing our CLAD algorithm that takes in the LFP signals as input and outputs infusion rates; and (3) a computer-controlled infusion pump that administers intravenous propofol. Our CLAD system controls an empirically-determined LFP marker of unconsciousness (MOU) at a user-prescribed target value by updating every 20 seconds the propofol infusion rate based on real-time processing of the LFP signal. The MOU is the instantaneous power in the 20 to 30 Hz band of the LFP spectrogram. Every cycle (duration{approx} 20 sec), our CLAD algorithm updates the MOU estimate and uses a robust optimal control strategy to adjust the propofol infusion rate based on the instantaneous error. This error is computed as the difference between the current and the user-prescribed target MOU values. Using neural recordings from multiple NHP anesthesia sessions, we first established that our chosen MOU signal was strongly correlated with propofol-induced decreased spiking activity which itself has been shown earlier to be associated with the level of unconsciousness in NHPs. Then we designed robust optimal control strategies that used subject-specific pharmacokinetic-pharmacodynamic models describing the MOU dynamics due to propofol infusion rate changes. Finally, we achieved safe and efficient closed-loop control of level of unconsciousness in 9 CLAD experiments involving 2 NHPs and 2 different 125 min long target MOU profiles with three target MOU changes within a given experiment. Our CLAD system performs stably, accurately and robustly across a total of 1125 min of closed-loop control. The CLAD performance measures, represented as median (25th percentile, 75th percentile), are 3.13 % (2.62%, 3.53%) for inaccuracy, 0.54 %(-0.31%, 0.89%) for bias, -0.02%/min (-0.06%/min, 0.00%/min) for divergence, and 3% (2.49%, 3.59%) for wobble. These performance measures were comparable or superior to previously reported CLAD performance measures from clinical studies (conducted outside USA) as well as rodent-based studies. The key innovations here are: (1) a pre-clinical NHP model for CLAD development and testing, (2) a neuroscience-informed LFP-based MOU for CLAD, (3) parsimonious, pharmacology-informed models to describe MOU dynamics under propofol infusion in rhesus macaques, (4) a novel numerical testing framework for propofol CLAD that incorporates a principled optimal robust control strategy for titrating propofol, and finally (5) experimental findings demonstrating the feasibility of stable, accurate and robust CLAD in the NHP model. Our NHP-based CLAD framework provides a principled pre-clinical research platform that can form the foundation for future clinical studies.

bioengineering↗

bioRxiv2021

Beyond dimension reduction: Stable electric fields emerge from and allow representational drift

It is known that the exact neurons maintaining a given memory (the neural ensemble) change from trial to trial. This raises the question of how the brain achieves stability in the face of this representational drift. Here, we demonstrate that this stability emerges at the level of the electric fields that arise from neural activity. We show that electric fields carry information about working memory content. The electric fields, in turn, can act as "guard rails" that funnel higher dimensional variable neural activity along stable lower dimensional routes. We obtained the latent space associated with each memory. We then confirmed the stability of the electric field by mapping the latent space to different cortical patches (that comprise a neural ensemble) and reconstructing information flow between patches. Stable electric fields can allow latent states to be transferred between brain areas, in accord with modern engram theory.

neuroscience↗