25th Annual Computational Neuroscience Meeting: CNS-2016

Abstract

A1 Functional advantages of cell-type heterogeneity in neural circuits

Tatyana O. Sharpee

A2 Mesoscopic modeling of propagating waves in visual cortex

Alain Destexhe

A3 Dynamics and biomarkers of mental disorders

Mitsuo Kawato

F1 Precise recruitment of spiking output at theta frequencies requires dendritic h-channels in multi-compartment models of oriens-lacunosum/moleculare hippocampal interneurons

Vladislav Sekulić, Frances K. Skinner

F2 Kernel methods in reconstruction of current sources from extracellular potentials for single cells and the whole brains

Daniel K. Wójcik, Chaitanya Chintaluri, Dorottya Cserpán, Zoltán Somogyvári

F3 The synchronized periods depend on intracellular transcriptional repression mechanisms in circadian clocks.

Jae Kyoung Kim, Zachary P. Kilpatrick, Matthew R. Bennett, Kresimir Josić

O1 Assessing irregularity and coordination of spiking-bursting rhythms in central pattern generators

Irene Elices, David Arroyo, Rafael Levi, Francisco B. Rodriguez, Pablo Varona

O2 Regulation of top-down processing by cortically-projecting parvalbumin positive neurons in basal forebrain

Eunjin Hwang, Bowon Kim, Hio-Been Han, Tae Kim, James T. McKenna, Ritchie E. Brown, Robert W. McCarley, Jee Hyun Choi

O3 Modeling auditory stream segregation, build-up and bistability

James Rankin, Pamela Osborn Popp, John Rinzel

O4 Strong competition between tonotopic neural ensembles explains pitch-related dynamics of auditory cortex evoked fields

Alejandro Tabas, André Rupp, Emili Balaguer-Ballester

O5 A simple model of retinal response to multi-electrode stimulation

Matias I. Maturana, David B. Grayden, Shaun L. Cloherty, Tatiana Kameneva, Michael R. Ibbotson, Hamish Meffin

O6 Noise correlations in V4 area correlate with behavioral performance in visual discrimination task

Veronika Koren, Timm Lochmann, Valentin Dragoi, Klaus Obermayer

O7 Input-location dependent gain modulation in cerebellar nucleus neurons

Maria Psarrou, Maria Schilstra, Neil Davey, Benjamin Torben-Nielsen, Volker Steuber

O8 Analytic solution of cable energy function for cortical axons and dendrites

Huiwen Ju, Jiao Yu, Michael L. Hines, Liang Chen, Yuguo Yu

O9 C. elegans interactome: interactive visualization of Caenorhabditis elegans worm neuronal network

Jimin Kim, Will Leahy, Eli Shlizerman

O10 Is the model any good? Objective criteria for computational neuroscience model selection

Justas Birgiolas, Richard C. Gerkin, Sharon M. Crook

O11 Cooperation and competition of gamma oscillation mechanisms

Atthaphon Viriyopase, Raoul-Martin Memmesheimer, Stan Gielen

O12 A discrete structure of the brain waves

Yuri Dabaghian, Justin DeVito, Luca Perotti

O13 Direction-specific silencing of the Drosophila gaze stabilization system

Anmo J. Kim, Lisa M. Fenk, Cheng Lyu, Gaby Maimon

O14 What does the fruit fly think about values? A model of olfactory associative learning

Chang Zhao, Yves Widmer, Simon Sprecher,Walter Senn

O15 Effects of ionic diffusion on power spectra of local field potentials (LFP)

Geir Halnes, Tuomo Mäki-Marttunen, Daniel Keller, Klas H. Pettersen,Ole A. Andreassen, Gaute T. Einevoll

O16 Large-scale cortical models towards understanding relationship between brain structure abnormalities and cognitive deficits

Yasunori Yamada

O17 Spatial coarse-graining the brain: origin of minicolumns

Moira L. Steyn-Ross, D. Alistair Steyn-Ross

O18 Modeling large-scale cortical networks with laminar structure

Jorge F. Mejias, John D. Murray, Henry Kennedy, Xiao-Jing Wang

O19 Information filtering by partial synchronous spikes in a neural population

Alexandra Kruscha, Jan Grewe, Jan Benda, Benjamin Lindner

O20 Decoding context-dependent olfactory valence in Drosophila

Laurent Badel, Kazumi Ohta, Yoshiko Tsuchimoto, Hokto Kazama

P1 Neural network as a scale-free network: the role of a hub

B. Kahng

P2 Hemodynamic responses to emotions and decisions using near-infrared spectroscopy optical imaging

Nicoladie D. Tam

P3 Phase space analysis of hemodynamic responses to intentional movement directions using functional near-infrared spectroscopy (fNIRS) optical imaging technique

Nicoladie D.Tam, Luca Pollonini, George Zouridakis

P4 Modeling jamming avoidance of weakly electric fish

Jaehyun Soh, DaeEun Kim

P5 Synergy and redundancy of retinal ganglion cells in prediction

Minsu Yoo, S. E. Palmer

P6 A neural field model with a third dimension representing cortical depth

Viviana Culmone, Ingo Bojak

P7 Network analysis of a probabilistic connectivity model of the Xenopus tadpole spinal cord

Andrea Ferrario, Robert Merrison-Hort, Roman Borisyuk

P8 The recognition dynamics in the brain

Chang Sub Kim

P9 Multivariate spike train analysis using a positive definite kernel

Taro Tezuka

P10 Synchronization of burst periods may govern slow brain dynamics during general anesthesia

Pangyu Joo

P11 The ionic basis of heterogeneity affects stochastic synchrony

Young-Ah Rho, Shawn D. Burton, G. Bard Ermentrout, Jaeseung Jeong, Nathaniel N. Urban

P12 Circular statistics of noise in spike trains with a periodic component

Petr Marsalek

P14 Representations of directions in EEG-BCI using Gaussian readouts

Hoon-Hee Kim, Seok-hyun Moon, Do-won Lee, Sung-beom Lee, Ji-yong Lee, Jaeseung Jeong

P15 Action selection and reinforcement learning in basal ganglia during reaching movements

Yaroslav I. Molkov, Khaldoun Hamade, Wondimu Teka, William H. Barnett, Taegyo Kim, Sergey Markin, Ilya A. Rybak

P17 Axon guidance: modeling axonal growth in T-Junction assay

Csaba Forro, Harald Dermutz, László Demkó, János Vörös

P19 Transient cell assembly networks encode persistent spatial memories

Yuri Dabaghian, Andrey Babichev

P20 Theory of population coupling and applications to describe high order correlations in large populations of interacting neurons

Haiping Huang

P21 Design of biologically-realistic simulations for motor control

Sergio Verduzco-Flores

P22 Towards understanding the functional impact of the behavioural variability of neurons

Filipa Dos Santos, Peter Andras

P23 Different oscillatory dynamics underlying gamma entrainment deficits in schizophrenia

Christoph Metzner, Achim Schweikard, Bartosz Zurowski

P24 Memory recall and spike frequency adaptation

James P. Roach, Leonard M. Sander, Michal R. Zochowski

P25 Stability of neural networks and memory consolidation preferentially occur near criticality

Quinton M. Skilling, Nicolette Ognjanovski, Sara J. Aton, Michal Zochowski

P26 Stochastic Oscillation in Self-Organized Critical States of Small Systems: Sensitive Resting State in Neural Systems

Sheng-Jun Wang, Guang Ouyang, Jing Guang, Mingsha Zhang, K. Y. Michael Wong, Changsong Zhou

P27 Neurofield: a C++ library for fast simulation of 2D neural field models

Peter A. Robinson, Paula Sanz-Leon, Peter M. Drysdale, Felix Fung, Romesh G. Abeysuriya, Chris J. Rennie, Xuelong Zhao

P28 Action-based grounding: Beyond encoding/decoding in neural code

Yoonsuck Choe, Huei-Fang Yang

P29 Neural computation in a dynamical system with multiple time scales

Yuanyuan Mi, Xiaohan Lin, Si Wu

P30 Maximum entropy models for 3D layouts of orientation selectivity

Joscha Liedtke, Manuel Schottdorf, Fred Wolf

P31 A behavioral assay for probing computations underlying curiosity in rodents

Yoriko Yamamura, Jeffery R. Wickens

P32 Using statistical sampling to balance error function contributions to optimization of conductance-based models

Timothy Rumbell, Julia Ramsey, Amy Reyes, Danel Draguljić, Patrick R. Hof, Jennifer Luebke, Christina M. Weaver

P33 Exploration and implementation of a self-growing and self-organizing neuron network building algorithm

Hu He, Xu Yang, Hailin Ma, Zhiheng Xu, Yuzhe Wang

P34 Disrupted resting state brain network in obese subjects: a data-driven graph theory analysis

Kwangyeol Baek, Laurel S. Morris, Prantik Kundu, Valerie Voon

P35 Dynamics of cooperative excitatory and inhibitory plasticity

Everton J. Agnes, Tim P. Vogels

P36 Frequency-dependent oscillatory signal gating in feed-forward networks of integrate-and-fire neurons

William F. Podlaski, Tim P. Vogels

P37 Phenomenological neural model for adaptation of neurons in area IT

Martin Giese, Pradeep Kuravi, Rufin Vogels

P38 ICGenealogy: towards a common topology of neuronal ion channel function and genealogy in model and experiment

Alexander Seeholzer, William Podlaski, Rajnish Ranjan, Tim Vogels

P39 Temporal input discrimination from the interaction between dynamic synapses and neural subthreshold oscillations

Joaquin J. Torres, Fabiano Baroni, Roberto Latorre, Pablo Varona

P40 Different roles for transient and sustained activity during active visual processing

Bart Gips, Eric Lowet, Mark J. Roberts, Peter de Weerd, Ole Jensen, Jan van der Eerden

P41 Scale-free functional networks of 2D Ising model are highly robust against structural defects: neuroscience implications

Abdorreza Goodarzinick, Mohammad D. Niry, Alireza Valizadeh

P42 High frequency neuron can facilitate propagation of signal in neural networks

Aref Pariz, Shervin S. Parsi, Alireza Valizadeh

P43 Investigating the effect of Alzheimer’s disease related amyloidopathy on gamma oscillations in the CA1 region of the hippocampus

Julia M. Warburton, Lucia Marucci, Francesco Tamagnini, Jon Brown, Krasimira Tsaneva-Atanasova

P44 Long-tailed distributions of inhibitory and excitatory weights in a balanced network with eSTDP and iSTDP

Florence I. Kleberg, Jochen Triesch

P45 Simulation of EMG recording from hand muscle due to TMS of motor cortex

Bahar Moezzi, Nicolangelo Iannella, Natalie Schaworonkow, Lukas Plogmacher, Mitchell R. Goldsworthy, Brenton Hordacre, Mark D. McDonnell, Michael C. Ridding, Jochen Triesch

P46 Structure and dynamics of axon network formed in primary cell culture

Martin Zapotocky, Daniel Smit, Coralie Fouquet, Alain Trembleau

P47 Efficient signal processing and sampling in random networks that generate variability

Sakyasingha Dasgupta, Isao Nishikawa, Kazuyuki Aihara, Taro Toyoizumi

P48 Modeling the effect of riluzole on bursting in respiratory neural networks

Daniel T. Robb, Nick Mellen, Natalia Toporikova

P49 Mapping relaxation training using effective connectivity analysis

Rongxiang Tang, Yi-Yuan Tang

P50 Modeling neuron oscillation of implicit sequence learning

Guangsheng Liang, Seth A. Kiser, James H. Howard, Jr., Yi-Yuan Tang

P51 The role of cerebellar short-term synaptic plasticity in the pathology and medication of downbeat nystagmus

Julia Goncharenko, Neil Davey, Maria Schilstra, Volker Steuber

P52 Nonlinear response of noisy neurons

Sergej O. Voronenko, Benjamin Lindner

P53 Behavioral embedding suggests multiple chaotic dimensions underlie C. elegans locomotion

Tosif Ahamed, Greg Stephens

P54 Fast and scalable spike sorting for large and dense multi-electrodes recordings

Pierre Yger, Baptiste Lefebvre, Giulia Lia Beatrice Spampinato, Elric Esposito, Marcel Stimberg et Olivier Marre

P55 Sufficient sampling rates for fast hand motion tracking

Hansol Choi, Min-Ho Song

P56 Linear readout of object manifolds

SueYeon Chung, Dan D. Lee, Haim Sompolinsky

P57 Differentiating models of intrinsic bursting and rhythm generation of the respiratory pre-Bötzinger complex using phase response curves

Ryan S. Phillips, Jeffrey Smith

P58 The effect of inhibitory cell network interactions during theta rhythms on extracellular field potentials in CA1 hippocampus

Alexandra Pierri Chatzikalymniou, Katie Ferguson, Frances K. Skinner

P59 Expansion recoding through sparse sampling in the cerebellar input layer speeds learning

N. Alex Cayco Gajic, Claudia Clopath, R. Angus Silver

P60 A set of curated cortical models at multiple scales on Open Source Brain

Padraig Gleeson, Boris Marin, Sadra Sadeh, Adrian Quintana, Matteo Cantarelli, Salvador Dura-Bernal, William W. Lytton, Andrew Davison, R. Angus Silver

P61 A synaptic story of dynamical information encoding in neural adaptation

Luozheng Li, Wenhao Zhang, Yuanyuan Mi, Dahui Wang, Si Wu

P62 Physical modeling of rule-observant rodent behavior

Youngjo Song, Sol Park, Ilhwan Choi, Jaeseung Jeong, Hee-sup Shin

P64 Predictive coding in area V4 and prefrontal cortex explains dynamic discrimination of partially occluded shapes

Hannah Choi, Anitha Pasupathy, Eric Shea-Brown

P65 Stability of FORCE learning on spiking and rate-based networks

Dongsung Huh, Terrence J. Sejnowski

P66 Stabilising STDP in striatal neurons for reliable fast state recognition in noisy environments

Simon M. Vogt, Arvind Kumar, Robert Schmidt

P67 Electrodiffusion in one- and two-compartment neuron models for characterizing cellular effects of electrical stimulation

Stephen Van Wert, Steven J. Schiff

P68 STDP improves speech recognition capabilities in spiking recurrent circuits parameterized via differential evolution Markov Chain Monte Carlo

Richard Veale, Matthias Scheutz

P69 Bidirectional transformation between dominant cortical neural activities and phase difference distributions

Sang Wan Lee

P70 Maturation of sensory networks through homeostatic structural plasticity

Júlia Gallinaro, Stefan Rotter

P71 Corticothalamic dynamics: structure, number of solutions and stability of steady-state solutions in the space of synaptic couplings

Paula Sanz-Leon, Peter A. Robinson

P72 Optogenetic versus electrical stimulation of the parkinsonian basal ganglia. Computational study

Leonid L. Rubchinsky, Chung Ching Cheung, Shivakeshavan Ratnadurai-Giridharan

P73 Exact spike-timing distribution reveals higher-order interactions of neurons

Safura Rashid Shomali, Majid Nili Ahmadabadi, Hideaki Shimazaki, S. Nader Rasuli

P74 Neural mechanism of visual perceptual learning using a multi-layered neural network

Xiaochen Zhao, Malte J. Rasch

P75 Inferring collective spiking dynamics from mostly unobserved systems

Jens Wilting, Viola Priesemann

P76 How to infer distributions in the brain from subsampled observations

Anna Levina, Viola Priesemann

P77 Influences of embedding and estimation strategies on the inferred memory of single spiking neurons

Lucas Rudelt, Joseph T. Lizier, Viola Priesemann

P78 A nearest-neighbours based estimator for transfer entropy between spike trains

Joseph T. Lizier, Richard E. Spinney, Mikail Rubinov, Michael Wibral, Viola Priesemann

P79 Active learning of psychometric functions with multinomial logistic models

Ji Hyun Bak, Jonathan Pillow

P81 Inferring low-dimensional network dynamics with variational latent Gaussian process

Yuan Zaho, Il Memming Park

P82 Computational investigation of energy landscapes in the resting state subcortical brain network

Jiyoung Kang, Hae-Jeong Park

P83 Local repulsive interaction between retinal ganglion cells can generate a consistent spatial periodicity of orientation map

Jaeson Jang, Se-Bum Paik

P84 Phase duration of bistable perception reveals intrinsic time scale of perceptual decision under noisy condition

Woochul Choi, Se-Bum Paik

P85 Feedforward convergence between retina and primary visual cortex can determine the structure of orientation map

Changju Lee, Jaeson Jang, Se-Bum Paik

P86 Computational method classifying neural network activity patterns for imaging data

Min Song, Hyeonsu Lee, Se-Bum Paik

P87 Symmetry of spike-timing-dependent-plasticity kernels regulates volatility of memory

Youngjin Park, Woochul Choi^, Se-Bum Paik

P88 Effects of time-periodic coupling strength on the first-spike latency dynamics of a scale-free network of stochastic Hodgkin-Huxley neurons

Ergin Yilmaz, Veli Baysal, Mahmut Ozer

P89 Spectral properties of spiking responses in V1 and V4 change within the trial and are highly relevant for behavioral performance

Veronika Koren, Klaus Obermayer

P90 Methods for building accurate models of individual neurons

Daniel Saska, Thomas Nowotny

P91 A full size mathematical model of the early olfactory system of honeybees

Ho Ka Chan, Alan Diamond, Thomas Nowotny

P92 Stimulation-induced tuning of ongoing oscillations in spiking neural networks

Christoph S. Herrmann, Micah M. Murray, Silvio Ionta, Axel Hutt, Jérémie Lefebvre

P93 Decision-specific sequences of neural activity in balanced random networks driven by structured sensory input

Philipp Weidel, Renato Duarte, Abigail Morrison

P94 Modulation of tuning induced by abrupt reduction of SST cell activity

Jung H. Lee, Ramakrishnan Iyer, Stefan Mihalas

P95 The functional role of VIP cell activation during locomotion

Jung H. Lee, Ramakrishnan Iyer, Christof Koch, Stefan Mihalas

P96 Stochastic inference with spiking neural networks

Mihai A. Petrovici, Luziwei Leng, Oliver Breitwieser, David Stöckel, Ilja Bytschok, Roman Martel, Johannes Bill, Johannes Schemmel, Karlheinz Meier

P97 Modeling orientation-selective electrical stimulation with retinal prostheses

Timothy B. Esler, Anthony N. Burkitt, David B. Grayden, Robert R. Kerr, Bahman Tahayori, Hamish Meffin

P98 Ion channel noise can explain firing correlation in auditory nerves

Bahar Moezzi, Nicolangelo Iannella, Mark D. McDonnell

P99 Limits of temporal encoding of thalamocortical inputs in a neocortical microcircuit

Max Nolte, Michael W. Reimann, Eilif Muller, Henry Markram

P100 On the representation of arm reaching movements: a computational model

Antonio Parziale, Rosa Senatore, Angelo Marcelli

P101 A computational model for investigating the role of cerebellum in acquisition and retention of motor behavior

Rosa Senatore, Antonio Parziale, Angelo Marcelli

P102 The emergence of semantic categories from a large-scale brain network of semantic knowledge

K. Skiker, M. Maouene

P103 Multiscale modeling of M1 multitarget pharmacotherapy for dystonia

Samuel A. Neymotin, Salvador Dura-Bernal, Alexandra Seidenstein, Peter Lakatos, Terence D. Sanger, William W. Lytton

P104 Effect of network size on computational capacity

Salvador Dura-Bernal, Rosemary J. Menzies, Campbell McLauchlan, Sacha J. van Albada, David J. Kedziora, Samuel Neymotin, William W. Lytton, Cliff C. Kerr

P105 NetPyNE: a Python package for NEURON to facilitate development and parallel simulation of biological neuronal networks

Salvador Dura-Bernal, Benjamin A. Suter, Samuel A. Neymotin, Cliff C. Kerr, Adrian Quintana, Padraig Gleeson, Gordon M. G. Shepherd, William W. Lytton

P107 Inter-areal and inter-regional inhomogeneity in co-axial anisotropy of Cortical Point Spread in human visual areas

Juhyoung Ryu, Sang-Hun Lee

P108 Two bayesian quanta of uncertainty explain the temporal dynamics of cortical activity in the non-sensory areas during bistable perception

Joonwon Lee, Sang-Hun Lee

P109 Optimal and suboptimal integration of sensory and value information in perceptual decision making

Hyang Jung Lee, Sang-Hun Lee

P110 A Bayesian algorithm for phoneme Perception and its neural implementation

Daeseob Lim, Sang-Hun Lee

P111 Complexity of EEG signals is reduced during unconsciousness induced by ketamine and propofol

Jisung Wang, Heonsoo Lee

P112 Self-organized criticality of neural avalanche in a neural model on complex networks

Nam Jung, Le Anh Quang, Seung Eun Maeng, Tae Ho Lee, Jae Woo Lee

P113 Dynamic alterations in connection topology of the hippocampal network during ictal-like epileptiform activity in an in vitro rat model

Chang-hyun Park, Sora Ahn, Jangsup Moon, Yun Seo Choi, Juhee Kim, Sang Beom Jun, Seungjun Lee, Hyang Woon Lee

P114 Computational model to replicate seizure suppression effect by electrical stimulation

Sora Ahn, Sumin Jo, Eunji Jun, Suin Yu, Hyang Woon Lee, Sang Beom Jun, Seungjun Lee

P115 Identifying excitatory and inhibitory synapses in neuronal networks from spike trains using sorted local transfer entropy

Felix Goetze, Pik-Yin Lai

P116 Neural network model for obstacle avoidance based on neuromorphic computational model of boundary vector cell and head direction cell

Seonghyun Kim, Jeehyun Kwag

P117 Dynamic gating of spike pattern propagation by Hebbian and anti-Hebbian spike timing-dependent plasticity in excitatory feedforward network model

Hyun Jae Jang, Jeehyun Kwag

P118 Inferring characteristics of input correlations of cells exhibiting up-down state transitions in the rat striatum

Marko Filipović, Ramon Reig, Ad Aertsen, Gilad Silberberg, Arvind Kumar

P119 Graph properties of the functional connected brain under the influence of Alzheimer’s disease

Claudia Bachmann, Simone Buttler, Heidi Jacobs, Kim Dillen, Gereon R. Fink, Juraj Kukolja, Abigail Morrison

P120 Learning sparse representations in the olfactory bulb

Daniel Kepple, Hamza Giaffar, Dima Rinberg, Steven Shea, Alex Koulakov

P121 Functional classification of homologous basal-ganglia networks

Jyotika Bahuguna,Tom Tetzlaff, Abigail Morrison, Arvind Kumar, Jeanette Hellgren Kotaleski

P122 Short term memory based on multistability

Tim Kunze, Andre Peterson, Thomas Knösche

P123 A physiologically plausible, computationally efficient model and simulation software for mammalian motor units

Minjung Kim, Hojeong Kim

P125 Decoding laser-induced somatosensory information from EEG

Ji Sung Park, Ji Won Yeon, Sung-Phil Kim

P126 Phase synchronization of alpha activity for EEG-based personal authentication

Jae-Hwan Kang, Chungho Lee, Sung-Phil Kim

P129 Investigating phase-lags in sEEG data using spatially distributed time delays in a large-scale brain network model

Andreas Spiegler, Spase Petkoski, Matias J. Palva, Viktor K. Jirsa

P130 Epileptic seizures in the unfolding of a codimension-3 singularity

Maria L. Saggio, Silvan F. Siep, Andreas Spiegler, William C. Stacey, Christophe Bernard, Viktor K. Jirsa

P131 Incremental dimensional exploratory reasoning under multi-dimensional environment

Oh-hyeon Choung, Yong Jeong

P132 A low-cost model of eye movements and memory in personal visual cognition

Yong-il Lee, Jaeseung Jeong

P133 Complex network analysis of structural connectome of autism spectrum disorder patients

Su Hyun Kim, Mir Jeong, Jaeseung Jeong

P134 Cognitive motives and the neural correlates underlying human social information transmission, gossip

Jeungmin Lee, Jaehyung Kwon, Jerald D. Kralik, Jaeseung Jeong

P135 EEG hyperscanning detects neural oscillation for the social interaction during the economic decision-making

Jaehwan Jahng, Dong-Uk Hwang, Jaeseung Jeong

P136 Detecting purchase decision based on hyperfrontality of the EEG

Jae-Hyung Kwon, Sang-Min Park, Jaeseung Jeong

P137 Vulnerability-based critical neurons, synapses, and pathways in the Caenorhabditis elegans connectome

Seongkyun Kim, Hyoungkyu Kim, Jerald D. Kralik, Jaeseung Jeong

P138 Motif analysis reveals functionally asymmetrical neurons in C. elegans

Pyeong Soo Kim, Seongkyun Kim, Hyoungkyu Kim, Jaeseung Jeong

P139 Computational approach to preference-based serial decision dynamics: do temporal discounting and working memory affect it?

Sangsup Yoon, Jaehyung Kwon, Sewoong Lim, Jaeseung Jeong

P141 Social stress induced neural network reconfiguration affects decision making and learning in zebrafish

Choongseok Park, Thomas Miller, Katie Clements, Sungwoo Ahn, Eoon Hye Ji, Fadi A. Issa

P142 Descriptive, generative, and hybrid approaches for neural connectivity inference from neural activity data

JeongHun Baek, Shigeyuki Oba, Junichiro Yoshimoto, Kenji Doya, Shin Ishii

P145 Divergent-convergent synaptic connectivities accelerate coding in multilayered sensory systems

Thiago S. Mosqueiro, Martin F. Strube-Bloss, Brian Smith, Ramon Huerta

P146 Swinging networks

Michal Hadrava, Jaroslav Hlinka

P147 Inferring dynamically relevant motifs from oscillatory stimuli: challenges, pitfalls, and solutions

Hannah Bos, Moritz Helias

P148 Spatiotemporal mapping of brain network dynamics during cognitive tasks using magnetoencephalography and deep learning

Charles M. Welzig, Zachary J. Harper

P149 Multiscale complexity analysis for the segmentation of MRI images

Won Sup Kim, In-Seob Shin, Hyeon-Man Baek, Seung Kee Han

P150 A neuro-computational model of emotional attention

René Richter, Julien Vitay, Frederick Beuth, Fred H. Hamker

P151 Multi-site delayed feedback stimulation in parkinsonian networks

Kelly Toppin, Yixin Guo

P152 Bistability in Hodgkin–Huxley-type equations

Tatiana Kameneva, Hamish Meffin, Anthony N. Burkitt, David B. Grayden

P153 Phase changes in postsynaptic spiking due to synaptic connectivity and short term plasticity: mathematical analysis of frequency dependency

Mark D. McDonnell, Bruce P. Graham

P154 Quantifying resilience patterns in brain networks: the importance of directionality

Penelope J. Kale, Leonardo L. Gollo

P155 Dynamics of rate-model networks with separate excitatory and inhibitory populations

Merav Stern, L. F. Abbott

P156 A model for multi-stable dynamics in action recognition modulated by integration of silhouette and shading cues

Leonid A. Fedorov, Martin A. Giese

P157 Spiking model for the interaction between action recognition and action execution

Mohammad Hovaidi Ardestani, Martin Giese

P158 Surprise-modulated belief update: how to learn within changing environments?

Mohammad Javad Faraji, Kerstin Preuschoff, Wulfram Gerstner

P159 A fast, stochastic and adaptive model of auditory nerve responses to cochlear implant stimulation

Margriet J. van Gendt, Jeroen J. Briaire, Randy K. Kalkman, Johan H. M. Frijns

P160 Quantitative comparison of graph theoretical measures of simulated and empirical functional brain networks

Won Hee Lee, Sophia Frangou

P161 Determining discriminative properties of fMRI signals in schizophrenia using highly comparative time-series analysis

Ben D. Fulcher, Patricia H. P. Tran, Alex Fornito

P162 Emergence of narrowband LFP oscillations from completely asynchronous activity during seizures and high-frequency oscillations

Stephen V. Gliske, William C. Stacey, Eugene Lim, Katherine A. Holman, Christian G. Fink

P163 Neuronal diversity in structure and function: cross-validation of anatomical and physiological classification of retinal ganglion cells in the mouse

Jinseop S. Kim, Shang Mu, Kevin L. Briggman, H. Sebastian Seung, the EyeWirers

P164 Analysis and modelling of transient firing rate changes in area MT in response to rapid stimulus feature changes

Detlef Wegener, Lisa Bohnenkamp, Udo A. Ernst

P165 Step-wise model fitting accounting for high-resolution spatial measurements: construction of a layer V pyramidal cell model with reduced morphology

Tuomo Mäki-Marttunen, Geir Halnes, Anna Devor, Christoph Metzner, Anders M. Dale, Ole A. Andreassen, Gaute T. Einevoll

P166 Contributions of schizophrenia-associated genes to neuron firing and cardiac pacemaking: a polygenic modeling approach

Tuomo Mäki-Marttunen, Glenn T. Lines, Andy Edwards, Aslak Tveito, Anders M. Dale, Gaute T. Einevoll, Ole A. Andreassen

P167 Local field potentials in a 4 × 4 mm² multi-layered network model

Espen Hagen, Johanna Senk, Sacha J. van Albada, Markus Diesmann

P168 A spiking network model explains multi-scale properties of cortical dynamics

Maximilian Schmidt, Rembrandt Bakker, Kelly Shen, Gleb Bezgin, Claus-Christian Hilgetag, Markus Diesmann, Sacha Jennifer van Albada

P169 Using joint weight-delay spike-timing dependent plasticity to find polychronous neuronal groups

Haoqi Sun, Olga Sourina, Guang-Bin Huang, Felix Klanner, Cornelia Denk

P170 Tensor decomposition reveals RSNs in simulated resting state fMRI

Katharina Glomb, Adrián Ponce-Alvarez, Matthieu Gilson, Petra Ritter, Gustavo Deco

P171 Getting in the groove: testing a new model-based method for comparing task-evoked vs resting-state activity in fMRI data on music listening

Matthieu Gilson, Maria AG Witek, Eric F. Clarke, Mads Hansen, Mikkel Wallentin, Gustavo Deco, Morten L. Kringelbach^, Peter Vuust

P172 STochastic engine for pathway simulation (STEPS) on massively parallel processors

Guido Klingbeil, Erik De Schutter

P173 Toolkit support for complex parallel spatial stochastic reaction–diffusion simulation in STEPS

Weiliang Chen, Erik De Schutter

P174 Modeling the generation and propagation of Purkinje cell dendritic spikes caused by parallel fiber synaptic input

Yunliang Zang, Erik De Schutter

P175 Dendritic morphology determines how dendrites are organized into functional subunits

Sungho Hong, Akira Takashima, Erik De Schutter

P176 A model of Ca²⁺/calmodulin-dependent protein kinase II activity in long term depression at Purkinje cells

Criseida Zamora, Andrew R. Gallimore, Erik De Schutter

P177 Reward-modulated learning of population-encoded vectors for insect-like navigation in embodied agents

Dennis Goldschmidt, Poramate Manoonpong, Sakyasingha Dasgupta

P178 Data-driven neural models part II: connectivity patterns of human seizures

Philippa J. Karoly, Dean R. Freestone, Daniel Soundry, Levin Kuhlmann, Liam Paninski, Mark Cook

P179 Data-driven neural models part I: state and parameter estimation

Dean R. Freestone, Philippa J. Karoly, Daniel Soundry, Levin Kuhlmann, Mark Cook

P180 Spectral and spatial information processing in human auditory streaming

Jaejin Lee, Yonatan I. Fishman, Yale E. Cohen

P181 A tuning curve for the global effects of local perturbations in neural activity: Mapping the systems-level susceptibility of the brain

Leonardo L. Gollo, James A. Roberts, Luca Cocchi

P182 Diverse homeostatic responses to visual deprivation mediated by neural ensembles

Yann Sweeney, Claudia Clopath

P183 Opto-EEG: a novel method for investigating functional connectome in mouse brain based on optogenetics and high density electroencephalography

Soohyun Lee, Woo-Sung Jung, Jee Hyun Choi

P184 Biphasic responses of frontal gamma network to repetitive sleep deprivation during REM sleep

Bowon Kim, Youngsoo Kim, Eunjin Hwang, Jee Hyun Choi

P185 Brain-state correlate and cortical connectivity for frontal gamma oscillations in top-down fashion assessed by auditory steady-state response

Younginha Jung, Eunjin Hwang, Yoon-Kyu Song, Jee Hyun Choi

P186 Neural field model of localized orientation selective activation in V1

James Rankin, Frédéric Chavane

P187 An oscillatory network model of Head direction and Grid cells using locomotor inputs

Karthik Soman, Vignesh Muralidharan, V. Srinivasa Chakravarthy

P188 A computational model of hippocampus inspired by the functional architecture of basal ganglia

Karthik Soman, Vignesh Muralidharan, V. Srinivasa Chakravarthy

P189 A computational architecture to model the microanatomy of the striatum and its functional properties

Sabyasachi Shivkumar, Vignesh Muralidharan, V. Srinivasa Chakravarthy

P190 A scalable cortico-basal ganglia model to understand the neural dynamics of targeted reaching

Vignesh Muralidharan, Alekhya Mandali, B. Pragathi Priyadharsini, Hima Mehta, V. Srinivasa Chakravarthy

P191 Emergence of radial orientation selectivity from synaptic plasticity

Catherine E. Davey, David B. Grayden, Anthony N. Burkitt

P192 How do hidden units shape effective connections between neurons?

Braden A. W. Brinkman, Tyler Kekona, Fred Rieke, Eric Shea-Brown, Michael Buice

P193 Characterization of neural firing in the presence of astrocyte-synapse signaling

Maurizio De Pittà, Hugues Berry, Nicolas Brunel

P194 Metastability of spatiotemporal patterns in a large-scale network model of brain dynamics

James A. Roberts, Leonardo L. Gollo, Michael Breakspear

P195 Comparison of three methods to quantify detection and discrimination capacity estimated from neural population recordings

Gary Marsat, Jordan Drew, Phillip D. Chapman, Kevin C. Daly, Samual P. Bradley

P196 Quantifying the constraints for independent evoked and spontaneous NMDA receptor mediated synaptic transmission at individual synapses

Sat Byul Seo, Jianzhong Su, Ege T. Kavalali, Justin Blackwell

P199 Gamma oscillation via adaptive exponential integrate-and-fire neurons

LieJune Shiau, Laure Buhry, Kanishka Basnayake

P200 Visual face representations during memory retrieval compared to perception

Sue-Hyun Lee, Brandon A. Levy, Chris I. Baker

P201 Top-down modulation of sequential activity within packets modeled using avalanche dynamics

Timothée Leleu, Kazuyuki Aihara

Q28 An auto-encoder network realizes sparse features under the influence of desynchronized vascular dynamics

Ryan T. Philips, Karishma Chhabria, V. Srinivasa Chakravarthy

Fig. 1

A Model schematic: tone inputs I_A and I_B elicit pulsatile responses in A1, which are pooled as inputs to a three-population competition network. Central unit AB encodes integrated, peripheral units A and B encode segregated. Mutual inhibition between units and recurrent excitation are incorporated with adaptation and noise. B A1 inputs show early initial adaptation, also if a pause is present. Build-up function shows proportion segregated increasing over time, here shown for three tone-frequency differences, DF, with no pause (dashed) or with a pause (solid curves). Time-snapshots from model (filled circles) agree with data (empty circles with SEM error bars, N = 8)

Fig. 2

N100 m predictions in comparison with available data [1, 2] for a range of pure tones (A) and HCTs (B)

Fig. 3

a Spike triggered covariance showing the full set of stimuli (black dots) projected onto the first two principle components. Stimuli causing a spike formed two clusters: net cathodic first pulses (blue) and net anodic first pulse (red). b Electrical receptive fields superimposed on the electrode array are shown for the cathodic first (blue) and anodic first clusters (red)

Fig. 4

A Visualization of C. elegans dynome, B communication diagram between the dynome and the layout, C snapshots of visualization of C. elegans during the PLM/AVB excitations (forward crawling)

Fig. 5

The average deviations of models and cell electrophysiology properties as measured in multiples of the 95 % CI bounds of experimental data means. Dashed line represents 1 CI bound threshold. Top rows show average deviations across all models for each cell property. Bottom rows show deviations across all cell properties for each model

Fig. 6

Oscillations in full and reduced networks of reciprocally coupled pyramidal cells and interneurons. A, B Illustrate topologies of reduced networks that generate “pure” ING and “pure” PING, respectively, while C highlights the topology of a “full” network that could in principle generate either ING or PING oscillations or mixtures of both. D, E Frequency of pure ING-rhythm generated by the reduced network in A (blue line), pure PING-rhythm generated by the reduced network in b (red line), and rhythms generated by the full network in C (green line) as a function of mean current to I-cells I_0,I and as function of mean current to E-cells I_0,E, respectively. D Results for networks with type-I interneurons while E shows results for networks with type-II interneurons. Pyramidal cells are modeled as type-I Hodgkin–Huxley neurons

Fig. 7

The amplitudes of saccade-related potentials (SRPs) to HS and VS cells are strongly correlated with each cell’s visual sensitivity to rightward yaw motion stimuli. A Experimental apparatus. B Maximal-intensity z-projections of the lobula plate to visualize HS- or VS-cell neurites that are marked by a GAL4 enhancer trap line. C, D The amplitude of saccade-related potentials (SRPs) were inversely correlated with visual responses, when measured under rightward yaw motion stimuli, but not under clockwise roll motion stimuli. Each sample point corresponds to each cell type. Error bars indicate SEM

Fig. 8

Power spectrum of ECS potential in a simulation including ECS diffusion (blue line) and a simulation without ECS diffusion (red line). Units for frequency and power are Hz and mV²/Hz, respectively

Fig. 9

Responses to input to the left V1 in the two cortical models with normal/abnormal structural connectivity. A Average firing rates. B–D Cortical regions and cortical areas that significantly responded to the input

Fig. 10

Jamming avoidance response

Fig. 11

Predictive information in the retinal response is coded for independently. Red the mutual information between the binary population firing patterns at times t and t + Δt, for 1000 randomly selected groups of 5 cells from our 31-cell population. Time is binned in 16.67 ms bins, and the (rare) occurrence of two spikes in a bin is recorded as a ‘1’. Blue the sum of the mutual information between a single cell response at time t and the future response of the group at time t + Δt. Error bars indicate the standard error of the mean across groups. All information quantities are corrected for finite-size effects using quadratic extrapolation [3]

Fig. 12

A The 3D-NFM adds a dendritic dimension to the 2D one [1]. One single macrocolumn has inhibitory (I) and excitatory (E) subpopulations. B (Top) Discretization of the dendrite. (Bottom) Equilibrium membrane potential along the dendrite for two different synaptic inputs. C PSDs of h_e for the 2D- and 3D-NFM. Increasing the synaptic input recovers the lost alpha rhythm

Fig. 13

A The convoluted signal with different ATP recovery rates (J_ATP) and relative connection strengths (C). B Standard deviation of the convoluted signals

Fig. 14

Comparison of circular probability density functions of sine and beta density. A Beta density with parameters a = b = 3.3818, matches closely that of the sine function, used as a probability density function (PDF). Beta density with parameters a = b = 3 solid line, is matched by sine function y = 1.05 − 1.1 cos(2π x/1.1). B Cumulative distribution function (CDF) is shown for these densities together with the difference between the two CDFs multiplied by 100 to visualize the comparison of the two distributions. C For testing different vector strengths we use uniform distributions with pre-set vector strengths (ρ = 0.8, 0.5 and 0.08)

Fig. 15

Design of recurrent neural networks and readouts

Fig. 16

A The T-junction assay with an entry angle of 20°. The axon is expected to prefer a right-turn at this angle. B A simple model is constructed where the direction of growth of the axon is proportional to area (red) it can explore

Fig. 17

The time distances between the first and second spikes of the simulated PD neurons as a function of the g_K and g_CaT conductances of the neuron with variable conductances. A first spikes. B Second spikes. The PD neuron with fixed conductances had g_K = 1.5768 μS and g_CaT = 0.0225 μS

Fig. 18

Concept (A–E) and simulation results (F–H). A Four activities without any clear meaning. b Activities in A are V1 response to oriented lines. C Comparison of brain’s view of spikes (left; apparently intractable) and scientist’s view of spikes (right; decoding possible). D Visuomotor agent set up. E Invariance principle. F Ideal state(s)-action(a) mapping R(s, a) (a), learned R(s, a) (b: synthetic input), learned R(s, a) (c: natural input). G Input (a), initial gaze trajectory (b), and learned gaze trajectory (c). H Learned state-action mapping (a: unordered; b: reordered rows), and learned receptive fields (c: unordered; d: reordered as b) [1]

Fig. 19

Networks implement different computations. A Persistent activity; network can sustain activity after removing stimulus. B Adaptation; network activity attenuates to background level given continuous stimulus. C Anticipative tracking; D network response leads moving stimulus in a certain speed

Fig. 20

A Three-dimensional orientation domains with columnar correlation length of Λ. B String singularities of orientation domains in A. Typical scale of cats Λ ≈ 1 mm

Fig. 21

Schematic of the sound stimuli used in all conditions, and the predicted reward for each

Fig. 22

A–C Membrane potential of the synthetic target (black), and of randomly chosen members of the final population (colors, overlaid almost exactly), from three validation studies. Optimized 10 and 23 parameters in A–C respectively. D–F Parameter values used to generate synthetic data (black lines), and mean ± standard deviation of values recovered in the searches (colored circles), normalized to the range used in the optimization

Fig. 23

Neuron network generated by our algorithm

Fig. 24

A Disrupted resting state brain network in obese subjects. B Global network properties network-based statistics

Fig. 25

A Schematics representing the neuronal network. A group of 2000 excitatory neurons and 500 inhibitory neurons are recurrently connected with sparse connectivity and the excitatory neurons receive random input from an external pool of neurons. B Excitatory neurons’ mean firing-rate (top), mean excitatory weight onto excitatory neurons (middle) and mean inhibitory weight onto excitatory neurons (connections marked as plastic in A). Simulation of the neuronal network with a spike-based inhibitory learning rule is represented by green lines (STDP) while simulation with our novel spike-timing- and current-dependent learning rule is shown in yellow (STCDP). The dashed lines represent the fixed points imposed by the excitatory (high) and inhibitory (low) learning rules. The low fixed point only exists for the inhibitory STDP model (simulation represented by the green lines)

Fig. 26

Simulation results. A Decay of neural activity for multiple repetitions of the same stimulus. B Experiment adapting with effective and ineffective stimuli. C Dependence of the PSTH on adaptor duration and unadapted response (black)

Fig. 27

A Visualizations available on the web-resource [2] for model browsing. B Schematic of upload and evaluation. Both experimental current traces and mod files can be uploaded to our servers, where they are scored and compared to all models currently in the database. C Exemplary result of automated comparison: Current traces (recorded from “Ramp” and “Activation” voltage clamp protocols) of the uploaded model (red) together with mean (1st, 2nd, 3rd, 4th) and individual (gray) traces of the four most similar clusters of channel models in the database

Fig. 28

A Time–frequency representation of local field potential (LFP) locked to a microsaccade (MS) recorded in primate V1. B Time–frequency representation of simulated LFP. C Schematic representation of the model network illustrating input (injection current), recurrent connection pattern and output (spike trains). D The input to the neurons is best reflected in the simulated spike trains (output) during phase I, quantified by mutual information (MI). E Recurrent connection pattern is best reflected in the output during phase II

Fig. 29

Relevant parameters of functional network of 2D Ising model at critical point versus fraction of defect to the structural cells. A Power-law exponent of degree-distribution, B small-worldness measure, C average degree

Fig. 30

Inhomogeneity of input current on host network, increases the response of network. A, B Response of networks of neurons and chain of neurons, for different inhomogeneity on host network and host neuron, respectively

Fig. 31

Comparison of simulated and experimental EMG during A rest, B 10 % maximum voluntary contraction

Fig. 32

A Schematic illustrations of the two balanced QIF networks models considered in the present study. The left network consists of strongly coupled neurons without noise, while the right network consists of weak coupling among neurons with noisy input. B Nearly identical rate autocorrelation functions in the two networks. The red line (C ₀) represents the value of the autocorrelation at time 0 and cyan line (C _∞) is the value of auto-correlation function in the limit of large t. C Change in spiking probability for different network connectivity strengths $(\tilde{g}),$ after being stimulated by a brief input at time t = 0

Fig. 33

Summary of experiment on the effect of riluzole on the dependence of burst frequency on potassium concentration. Without riluzole (left), the frequency increases steadily with increasing potassium concentration. With riluzole present (right), the frequency remains essentially constant with increasing potassium concentration

Fig. 34

Nonlinear modulation of the firing rate by a cosine signal. A Signal, B subthreshold voltage, C rasterplot, D The time-dependent firing rate (red, noisy trace) is significantly different from the linear theory (dashed line) but is accurately described by the second-order response (solid line)

Fig. 35

Phase space portrait and divergence of nearby trajectories. A The top panel shows the orthogonal relationship between the forward and reversal behaviors, while the bottom panel shows the transition from reversal to an omega turn in the phase space. To aid visualization color coding is done by radial distance from the origin. B Escape response visualized in the phase planes. When the worm is hit with a laser impulse, it makes a reversal, followed by an omega turn and then resumes forward crawling. Color map encodes time in frames. C Divergence curves for the three different attractors. Y-axis shows the exponential of the divergence between neighboring trajectories plotted on a semilog scale on the y axis, each curve corresponds to a single worm (n = 12). λ _L is estimated by calculating the slope of the linear region. Boxplots show the range of λ _L obtained from different animals

Fig. 36

Experimental design and result. A Marker confusion. Grey dots are markers. d ₁, d ₂ are the distances between markers, t _s is the sampling latency, v is speed of marker. Green lines show the markers, which identified as same. Left correct identification right example of marker confusion. B Experimental set up. Red dots are keys to press by the thumb and the little finger during repeats. C The probabilities of continuous marker identification

Fig. 37

Theoretical predictions (lines) and numerical simulation (markers) are shown. A1 Classification of line segments. (Solid) lines embedded in the margin, (dotted) lines touching the margin, (striped) interior lines .A2 Capacity α = P/N of a network N = 200 as a function of R (line length) with margins κ = 0, 0.5. A3 Fraction of configurations at capacity with κ = 0. (red) lines in the margin, (blue) touching the margin, (black) interior lines. B1 D2 balls, B2 capacity α = P/N for κ = 0 for large D = 50 and R $\propto$ D^−1/2 as a function of $\text{R}\sqrt{\text{D}}$. (Blue solid) α_D(0, R) compared with α₀($\text{R}\sqrt{\text{D}}$) (red square). (Inset) capacity α at κ = 0 for 0.35 ≤ R ≤ 20 and D = 20: (blue) theoretical α compared with approximate form (1 + R⁻²)/D (red dashed). C1 2D L1 balls. C2 Fraction of configurations as a function of radius R at capacity with κ = 0. (red) entire manifold embedded, (blue) touching margin at a single vertex, (gray) touching with two corners (one side), (purple) interior manifold

Fig. 38

Example of the spatial attenuation of the extracellular potential signal for a particular set of inhibitory connections. The temporal traces at two electrode locations are represented with blue and green dots accordingly. Average over the absolute maximum extracellular potential amplitudes is shown in 2D space. According to the schematic the rate of the extracellular signal spatial attenuation generated by the pyramidal cell is approximately 400 μ

Fig. 39

Firing rates, synaptic efficacy and cross-correlation change during the adaptation. A The time course of firing rates and the averaged synaptic efficacy of the network during the adaptation. $〈\text{ux}〉$ is temporally enhanced during the adaptation due to the STF, but in the long term, strong STD drives the synaptic efficacy to background level. Stimulation is during 0–1500 ms. B The enhancement of cross-correlation between neurons during the adaptation

Fig. 40

A Success rate of two model rats which shows rule between them (blue dots represent the left cue success rate of model rat1, orange plus represent the right cue success rate of model rat1, yellow cross represent the left cue success rate of model rat2, and purple line represent the right cue success rate of model rat2). B Simulation result (600 iterations). C Behavior experiment result (19 pairs)

Fig. 41

A Schematic of the V4-PFC network model. B Optimal representation of the shape-selective V4 responses as a function of occlusion level. C Neuronal responses with a noise projected onto the test shape-selective (unit 1)/non-selective (unit 2) V4 response plane, before (top) and after (bottom) the feedback inputs from PFC. Feedback inputs move the responses away from the unity line, improving shape discriminability under occlusion

Fig. 42

A Fitness evolution generation 0–4000 (first 3 principle components). Each color is a different word class, each line a different utterance token of the word. B Change in state space trajectory from STDP adaptation

Fig. 43

Computational framework for analyzing space–time cortical dynamics. T is an idempotent projection matrix

Fig. 44

Network connectivity before and after sensory stimulation. A, B. Connectivity matrix, pre- and post-synaptic neurons are sorted according to their preferred orientation (PO) and subdivided into groups. C, D. Mean output connectivity plotted against the difference between pre and post PO

Fig. 45

Three dimensional subsets of the 8D corticothalmic coupling space. A, B Regions with 1, 3 or 5 roots are enclosed by surfaces (blue, violet and yellow respectively). The sharp transition between zones of 1–3 roots along the v_ee axis (excitatory intracortical feedback) indicates the plane at which the total intracortical feedback (v_ee + v_ei) changes sign. The difference between A, B is the value of v_ei (inhibitory intracortical feedback). In a the probability of having multiple roots is lower than in B

Fig. 46

V1 representations of the stimuli and simulation results of the model. A V1 representations of the stimuli by using NCRF model. B Model performance increases with perceptual training. C Synaptic weight changes after perceptual learning

Fig. 47

Subsampling scaling in model and experiment. Left branching process model; right: experiments on developing cultures A Avalanche size counts f(s) from the full and the subsampled critical model; N: number of sampled neurons. B Under subsampling scaling, all f(s) collapse. C Collapse of subsampled avalanche-size distribution from the culture at the age of 21 days. D For subcritical models, the same scaling ansatz does not result in a collapse. E No collapse of f(s) from the culture at age 7 days

Fig. 48

Relative active information storage as a function of the time range of the past state for different estimators

Fig. 49

Example of active learning, simulated with a three-alternatives model on 1D stimulus. After each observation, the psychometric functions are estimated based on the accumulated data, and the next stimulus is chosen to maximize the expected information gain. The estimated psychometric functions (solid lines) quickly approach the true functions (dashed lines) through the adaptive and optimal choice of stimuli

Fig. 50

Local repulsive interaction develops a consistent interference between mosaics. A Moiré pattern of RGC. B Developmental model of RGC mosaic with local repulsive interaction between nearby cells. C Developed cell mosaic. D Autocorrelation of developed mosaics. e Approach between ON and OFF mosaics induces a gradual reinforcement of heterotypic interaction. F Angle alignment between mosaics (θ) is limited to low angles as mosaics approach (*: p < 0.05, Ranksum test, error bar: SE)

Fig. 51

Correlation between bistable perception and perception under ambiguous signal. A Racetrack stimulus. Rotational motion can be either illusory or ambiguous depending on coherence. B Example response of racetrack. Perceived motion can be bistable (top) or follows actual motion with response time (bottom). C Subjects’ (black) and model’s (red) phase duration and response time are highly correlated. D Double-well energy model to describe behavior during bistable perception and perceptual decision making task

Fig. 52

The simulation model for developmental mechanism of salt-and-pepper map by feedforward convergence between retina and V1. A Moiré interference between ON and OFF RGC mosaics. B Moiré interference with small and large alignment angles can generate various range of periodicity, S. C From rat RGC mosaics, both smooth and salt-and-pepper map can be developed. D Spatial distribution of preferred orientations of V1 cells by different convergence conditions; Smooth map model (high sampling ratio and large convergence range), Salt-and-pepper map model (low sampling ratio and short convergence range). E The structure of orientation map depending on convergence range and sampling ratio

Fig. 53

A novel index effectively describes different neural activity patterns obtained from imaging. A Neural activity obtained from optical imaging could be analyzed with appearance and propagation. B Appearance index of four distinct sample. C Propagation index of straight trajectory (top) and curved trajectory (bottom). D Propagation Index of non-dispersive sample (top) and dispersive sample (bottom)

Fig. 54

Different learning rules reproduce volatile/nonvolatile memory system. A Spike timing dependent plasticity. B Memory decaying properties of different learning rule. Poisson spikes are given for 1000 s to simulate decaying environment. C, D Multiple patterns was given to the system every 200 s. C Memory performance of each pattern in AS memory system. D Memory performance of each pattern in SS memory system

Fig. 55

The statistics of the first-spike occurrence times (amplitude of TPCS ε₀ = 0.2, cell size S = 100 μm², frequency of suprathreshold signal f = 20 Hz and amplitude of it A = 4μA/cm²). A Mean latency of the network, B jitter of the network

Fig. 56

One-spike burster and estimated model as in Table 1

Fig. 57

Experimental ORN responses to stimulus measured by electro-antennogram recordings in [5] (top, black line) is qualitative similar to the average normalized ORN responses to stimulus (1-hexanol at concentration 0.1 M) predicted by our model (bottom)

Fig. 58

A Simulation geometry showing the four modeled layers: insulator (glass), vitreous, NFL, and GCL. Distance from membrane threshold in mV for B parallel axons in a plane in the NFL and C perpendicular axon initial segments in a plane in the GCL, when stimulated with a 300 µs biphasic pulse with electrode-retina separation of 400 µm. Dotted contour marks the threshold level

Fig. 59

A Mean spike-timing reliability (similar correlation-based measure as in [3], but with firing rate adaption). The reliability of the VPM input is 0.55. B Mean probability of firing within 2–12 ms after the initial input VPM spike in each trial. C Mean ratio of spikes occurring within 2–12 ms after a VPM spike, out of all spikes. Mean of 30 (L3/4 excitatory), 50 (L5/6 exc.), 40 (L3/4 inhibitory) and 30 cells (L5/6 inh.) respectively

Fig. 60

Stimuli and fMRI results. A The snapshot of traveling Gabors are shown for the four different conditions. The black arrows represent a moving direction of wedge or ring. B Significant (yellow, t test p < 0.001) coaxial anisotropy in all subjects. C Coaxial anisotropy across visual areas (V1, V2, V3)

Fig. 61

Bayesian estimation to predict BOLD dynamics around switch. A Bayesian inference model of iteratively updated prior, input likelihood, and combined posterior. B (Upper) Uncertainty-driven BOLD estimated from Bayesian model locked to transition under different duration conditions. Time-series is built purely from real behavior history. (Lower) Average % BOLD signal of ACC region in 8 subjects (14 sessions)

Fig. 62

Interaction between tuning curve and prior in population tuning. A Bias map for combinations of tuning curve width and prior sharpness. Negative bias means that perception of a near-/ba/stimulus was biased toward/ba/, namely categorical perception. Three inset plots on ordinate and abscissa show the cases of the lowest/median/highest concentration parameters of tuning curve and prior peak, respectively. B Discrimination difference map. Negative number indicates that between-phoneme condition outperformed near-/ba/condition. C Tuning curves of population neurons that was marked as green squares in A, B. Location of tuning centers were marked as dots for 60 neurons, and tuning curves of 30 out of those 60 neurons were drawn below

Fig. 63

A SE and FC values of Fp1 channel for three different states, which are wakeful, ketamine-induced, and propofol-induced states, are averaged over subjects (n = 29 for ketamine-induced, n = 20 for propofol-induced and n = 49 for wakeful states). Error bars represent standard errors. Wakeful state has the intermediate SE value between ones of two other states. For FC value, however, wakeful state has the highest one and both anesthetized states have smaller ones, forming a concave relationship between three states. B Each dot manifests averaged SE and FC values of one subject over 23 10 s-long epochs overlapping 5 s each other. Error bars here also indicate standard errors. Most wakeful states have higher FC values compared to ones of unconscious states and have intermediate SE values. Ketamine-induced states are mainly located in the lower right part when propofol-induced states are clustered at the lower left part of the area in SE–FC plot. C FC values of EEG signals from the whole brain area, covering pre-frontal, frontal, temporal, and parietal regions, significantly decreased during ketamine-induced loss of consciousness (p < 0.0001). D All FC values of the signals from different regions were also significantly reduced during propofol-induced unconscious state (p < 0.0001)

Fig. 64

Distribution of avalanche size for LHG model on the fully-connected network. The distribution function of avalanche size shows the power law with exponent −1.57

Fig. 65

A Microelectrodes placed to cover both entorhinal cortex and hippocampus. B Temporal changes in global and local efficiency around the time of ictal-like epileptiform activity. The red vertical line indicates the initiation of ictal-like events and the time series in blue displays recorded field potentials

Fig. 66

Recording data (A) and simulation results (B) of SLE suppression effect by electrical stimulation

Fig. 67

Sparse incomplete representations (SIR). In our previously formulated model of the main olfactory bulb network [1], MCs receive inputs from receptor neurons in the glomeruli (black circles) and interact with GCs through dendrodendritic synapses. GCs build representations of MC glomerular inputs (red arrows). The representations are contained in the inhibitory inputs returned by the GCs to the MCs (blue arrows). Because GCs inhibit each other through second-order inhibitory interactions, only a few GCs respond to an odorant (full blue circles with a dendrite shown). The vast majority of GCs do not change their firing rate in response to an odorant (empty circles). Thus, the responses of GCs are sparse. Because some MCs manage to retain the responses to odorants, the representation by GCs is called incomplete. According to this model, MCs transmit to higher areas the errors in the GC representation

Fig. 68

A Response of pyramidal cells to transient input to excitatory interneurons shows different modes. B Depending on intensity and duration of the stimulus

Fig. 69

Each node represents stimulation type and each edge means classification error rate. Length of edge shows similarity between a pair of stimulation

Fig. 70

Overall characteristics of alpha phase synchronization for PA. A The upper triangle of association matrix indicates the PLV calculated by grand mean phase coherence. The lower triangle of association matrix indicates the its CI values in pairs. B Topographical connections with the 12 top-ranked CI from the lower triangle of a

Fig. 71

A Model with local (left) and long-range connections (right). B, C Averaged tract lengths and weights from 4 connectomes B Joint distribution, and C histogram of weighted lengths for intra- and inter-hemispheric links. D Sketch of the spatial delay structure. E Phase-lag distributions (top) and phase-lags between areas (bottom rows)

Fig. 72

Multidimensional decision making task design and model comparison. A Multidimensional decision making task schematics. B Systemic structure of the task. C The result of model comparison, proposed model has significantly high accuracy on prediction. D, E The models’ prediction accuracy of proposed model (D) and naïve model (E)

Fig. 73

Low-cost eye movement tracker using front cameras on the each devices

Fig. 74

A Brain synchrony analyses. Intra-brain and inter-brain phase synchronies in alpha band [0.5, 1] s. Links between electrodes means that the phase activities there are synchronized. All synchronies here were higher in FF groups than in FB groups (gray line). Blue line denotes the synchronies that were higher in CC epochs compared with DD epochs of FF groups (CC > DD) whereas red line denotes the synchronies that were higher in DD epochs compared with CC epochs of FF groups (CC < DD). Intra-brain synchronies are drawn in both brains and only one of each pair of inter-brain synchronies are drawn. Significant level was at p < 0.05, Bonferroni corrected. B Magnitudes of phase synchronies that showed significant strategical differences. These correspond to the links depicted as blue and red line in A. * p < 0.05; ** p < 0.01, Bonferroni corrected

Fig. 75

Three critical pathways emerged from the results. A For V _B they were: (1) AVA-based; (2) PVP-based; and (3) RMD → OLL. B Two of these pathways were again implicated for V _E: the AVA-based and the PVP-based pathways

Fig. 76

A Comparison of the prescreening accuracy in terms of ROC-AUC value (higher is better accuracy). B Comparison of the computation time. ‘GFAM10 [1]’, regarded as the original method, denotes generative functional additive model which is extended version of generalized linear model. ‘Correlation-GFAM10’ denotes a hybrid approach which performs the Pearson correlation for prescreening and then performs GFAM10

Fig. 77

Early discrimination of stimulus in the MBs. A Recordings of PNs and MBONs activities from untrained honey bees to odor stimulation. At t = 0 s (green bar), an odor stimulation is presented. B Connectionist blueprint of the MBs, emphasizing synapses and population size. Note the divergence present between PN and KC layers, followed by convergence onto MBONs. C We reproduced in silico the early response (blue bar) of MBONs in the vertical lobe with respect to the PNs (orange bar) using spiking neuron networks. D Time differences in our simulations for each experiment repetition. Difference in response time between is on average 50 ms (one tail Mann–Whitney test, p < 0.025)

Fig. 78

Visualisations of processing and deep learning stages. A Wavelet decomposition across bands of interest. B Progression of image-encoded oscillatory synchronization in BA10. C Deep learning network improvement across training epochs. D One trained network layer displaying parcel dynamics that mediate classification

Fig. 79

A An MRI image from the BrainWeb [1] and the result of segmentation into five clusters. B An MRI image with 7 % noise added and the result of segmentation into five clusters

Fig. 80

A The network model. Plus symbol indicates excitatory connection, and minus symbol indicates inhibitory connection. LFP is computed from GPi and STN populations separately. These two LFP signals are used as the source of the five MDFS stimulations (dashed-arrows): STN-to-STN, STN-to-GPi, GPi-to-STN, GPi-to-GPe and GPi-to-GPi, shown by arrow labeled 1–5. B Error index values for 80 different model TC neurons in an intermittent network. Comparison of MDFS among different stimulation targets using either STN or GPi LFP signal. Whisker plots show mean (red line), 25–75 percentile range (blue box), 95 % confidence interval (black lines) and outliers (red plus signs)

Fig. 81

Response of a modeled neuron for different initial conditions, V(0). Sodium, three types of potassium, calcium, and hyperpolarisation-activated currents are included in the model

Fig. 82

The phase of post-synaptic firing in response to frequency-modulated inhomogenous Poisson pre-synaptic spike trains depends on the configuration of input synapses. A The connectivity involves M independent pre-synaptic neurons each with NM synaptic release sites at a post-synaptic neuron. Vesicles are released probabilistically when activated by a pre-synaptic action-potential, if one is available at that site. The post-synaptic neuron depolarizes and produces action potentials after arrival of neurotransmitter according to standard models. B The phase lead of output spiking relative to periodic modulation at frequency f, for M active-zones is both frequency-dependent and configuration-dependent. The figure is from simulations but we also derived the same result mathematically

Fig. 83

A Model architecture with 2D neural field that receives input from two hierarchical path-ways. B Response traces of MP neurons for silhouette stimulus without shading during a 200 s simulation. C Corresponding average response times of the output neurons. D Response times for shaded stimulus

Fig. 84

A Model architecture consisting of two coupled neural fields, implemented with biophysically realistic neurons. B Psychophysical results from [1] showing the dependence of the detectability of visual point-light stimuli in dependence of the delay between a visually observed and the concurrently executed action. C Simulated detectability derived from the model for the same experimental conditions

Fig. 85

Relative error (RE) in percentage between graph theoretical measures of simulated FC versus empirical FC for the entire (1–100 %) and selected range of connection densities (37–50 %). Bars and error bars correspond respectively to the averages and standard deviations across the ten RE values. E _glob global efficiency, E _loc local efficiency, CC clustering coefficient, L characteristic path length, EC eigenvector centrality, PC participation coefficient, SW: small-worldness, ${\text{R}}_{\text{t}}^{\text{c}}$ and ${\text{R}}_{\text{t}}^{\text{g}}$ represent resilience to targeted attack in the size of largest connected component and global efficiency, respectively, ${\text{R}}_{\text{r}}^{\text{c}}$ and ${\text{R}}_{\text{r}}^{\text{g}}$ represent resilience to random failure in the size of largest connected component and global efficiency, respectively

Fig. 86

Normalized energy spectra and voltage traces resulting from asynchronous neural activity. A, B Results of superimposed, asynchronous action potential waveforms for quasi-periodic frequencies of 100 Hz (A) and 200 Hz (B). C, D Results of superimposed, asynchronous postsynaptic potential waveforms for quasi-periodic frequencies of 100 Hz (A) and 200 Hz (B). Gray dashed lines represent energy spectra that would result from Poisson process spike trains convolved with AP/PSP waveforms

Fig. 87

A, B Fits to motion onset responses to estimate each neuron’s kinetics. C Experimentally estimated MT transients to positive and negative speed changes of various magnitude. D Transient response amplitudes as derived from the model. E, F Relation between transient and sustained MT responses as a function of speed change magnitude as estimated experimentally (E) and by the model (F)

Fig. 88

Comparison of model with reduced (red) morphology to the model with full (blue) morphology. The y-axis shows the membrane potential at soma (solid) and apical dendrite (dashed) as a response to a somatic 200-ms DC pulse

Fig. 89

A Instantaneous spiking and LFP in a 4-layer network model covering 4 × 4 mm² at realistic cell and synapse density with distance-dependent connectivity. B Pairwise correlations between spike trains of exc. (E) and inh. (I) layer 5 neurons as function of distance (red: E–E, blue: I–I, black: E–I). C Distance-dependent LFP correlation computed for a 10 × 10 electrode grid in layer 5 (0.4 mm between contacts)

Fig. 90

A The spike raster plot showing 0.6 s of simulation. The vertical axis shows neuron index. Neurons from index 1 to 100 receive structured inputs. Colored spikes refer to PNGs founded by the same colored spikes of readout neurons (above the dash line), where the letter-marked ones are shown in other panels. B A recovered PNG with the predicted spike timing (receptive field) and the actual spikes. The numbers are neuron indices. C The same PNG in B but activated at another time. D, E Another PNG

Fig. 91

Example estimation of a seizure recording. A Sixteen channel electrocortiography (ECoG) of seizure (red lines indicate start and end points). B The ECoG channels are modelled as cortical regions, each with three coupled populations. C–G Estimation results of coupling strength (proportional to color) between neural populations for 16 cortical regions (vertical axis), over the time span of the seizure (horizontal axis)

Fig. 92

Data-driven model estimation. A The basic unit of a neural model is described by the mean membrane potential, v_n(t), of a neural ensemble and synaptic inputs. B, C Pre-synaptic firing rates are convolved with the excitatory/inhibitory kernel to generate membrane potential fluctuations. D The resulting membrane potential is converted to an output firing rate via a sigmoidal transform. E Electrical recording of a seizure. F Estimated gain parameters during seizure

Fig. 93

Changes in functional connectivity with respect to baseline after inhibitory stimulation as a function of cortical weight of the structural connectivity matrix. Red line: mean uniform bins curve smoothed; dashed line: mean weight

Fig. 94

Propagation patterns of optical stimulation at each frequency in thalamocotical circuit. Beta frequency stimulus propagated S1–M1 strongly, but gamma frequency case, contralateral propagation is dominant. Blue bars indicate optical stimulus in left VPM

Fig. 95

The pairs with statistically significantly increased (red) or decreased (blue) PSI of gamma oscillations (Student t test, p < 0.05). Only the pairs from the prefrontal cortex were depicted here. SD and R stand for sleep deprivation and recovery days, respectively

Fig. 96

A The model architecture. B Hexagonal firing field of a single neuron in the outer layer of the model while remaping its response on the visual space

Fig. 97

The model architecture (A) used to simulate the water maze task indicating the notion of a direct and an indirect pathway. The value function (B) developed after training the agent for 10 trials, the value peaks near to the platform. The escape latency (C) through trials shows that the agent has learnt the task with increased hippocampal dependence in the earlier stages and cortical dependence in the later stages of learning. The spectrogram (D) of the activity of CA3 as the function of time shows desynchronization while active exploration of the maze (0–45 s) and synchronized activity upon reaching the platform (45–65 s)

Fig. 98

A Centre surround mapping in striosomes and matrisomes. The striosomes highlighted by red, map the states and the matrisomes highlighted by green, map the actions. B Value function map in the multiple context setting where the reward is present at the top left corner and the bottom right corner. C Switching of the modules based on the environmental contexts. The reward changes every 1000 episodes and the corresponding change in module with episode is shown

Fig. 99

The model architecture (A) with the different modules aiding in reaching movements. The comparison of controls and PD’s approach to a target (B) and the appearance of PD symtoms including tremor and rigidity as a function of distance to the target

Fig. 100

A Self-history filters (diagonal) and directed coupling filters between neurons (off-diagonal) in the full 3-neuron network. Neuron 1 is excitatory and its couplings to the other neurons are strictly positive. Neurons 2 and 3 are inhibitory and make strictly negative couplings to other neurons. There is no coupling from neuron 2–3. B Effective self-history filters (diagonals) and coupling filters (off-diagonals) when neuron 3 is hidden. The bottom row is unaltered because neuron 2 makes no coupling to neuron 3. The filters in the top row are changed due to the influence of signals neuron 1 sends to itself through neuron 3 and to neuron 2 through neuron 3. Although neuron 1’s true self-history filter and coupling from neuron 2 are negative the effective filters change sign. C The effective self-history filter of neuron 1 when both neurons 2 and 3 are hidden. Times and filter amplitudes are given in arbitrary units (a.u.)

Fig. 101

Large-scale wave patterns for strong coupling, showing four time snapshots for a traveling wave (top), a spiral wave (middle), and a sink pattern (bottom). Warmer colors denote higher amplitudes

Fig. 102

In small synapses (200 nm × 200 nm), Ratios of maximum NMDA receptor opening probabilities as functions of receptor distance for different release speed (slow, 2 nm fusion pore—triangle, regular, 10 nm fusion pore—asterisk, and instantaneous—circle) of glutamate vesicle release. The open probabilities were calculated by the kinetics equation, when glutamates are released above the center location

Fig. 103

A1–A4 Network structure. Neurons of the first and second stored pattern are represented by colors ranging from blue to green and yellow to red, respectively. Effective synaptic connections can be calculated and are shown by colored segments. B1–B4 Cross-correlograms (CCG) of single neuron activity with the summed activity of other neurons (see [1]). C1–C4 Center of mass of CCGs, noted μ_CC

Fig. 104

A Auto-encoder NN coupled to the VN. B, C Depict desynchronized (ɛ = 1) and synchronized (ɛ = 0) states of VN respectively. The corresponding output weight patterns learnt by the auto-encoder, driven by the VN, trained on bar pattern data (B1, C1) and MNIST data (B2, C2)