Skip Navigation

CNS*2017 Workshop

Location: Antwerp, Belgium

CNS*2017 Workshop 

Cortical Function: Towards Understanding and Developing Integrative Theories 

July 20, 2017 

26th Annual Computational Neuroscience Meeting (CNS*2017)
Antwerp, Belgium.

Brief Description: Understanding how our brain computes and analyses sensory inputs from our external environment whilst enabling us to experience such rich and varied mental lives is one of the great scientific challenges of the 21st Century. Recent advances have uncovered much about the cerebral cortex, with its 2-4mm thick sheet of neurons having a consistent anatomical structure consisting of six well-characterised layers and network connectivity. This workshop aims to look at what progress has been made in understanding how the cortex functions and what general integrative principles underlie how it works and enable capabilities as diverse as sensory perception, control of voluntary motor activity and high-level cognitive functions.


  • Hamish Meffin, National Vision Research Institute, and Department of Optometry & Visual Science, The University of Melbourne, 
  • Anthony Burkitt, Department of Biomedical Engineering, The University of Melbourne,


14:00–14:45 Marcus Diesmann (Institute of Neuroscience and Medicine, Research Centre Jülich , Germany)

  • A brain-scale model of macaque visual cortex at cellular and synaptic resolution

14:45–15:30 Jorge Mejias (Center for Neural Science , New York University, USA)

  • Large-scale models of cortical dynamics: neural communication and cognitive computations

15:30–16:00 Coffee Break

16:00-16:45 Subutai Ahmad (VP Research Numenta, USA)

  •  Why the Neocortex Has Layers and Columns, a Theory of Learning 3D Models of the World

16:45-17:30 Stefan Mihalas (Allen Institute for Brain Science, USA)

  • Cortical circuits implement optimal context intefration and its gating


Abstract and Biographies

14:00–14:45 Marcus Diesmann (Institute of Neuroscience and Medicine, Research Centre Jülich , Germany)
A brain-scale model of macaque visual cortex at cellular and synaptic resolution 

Abstract: The cortical microcircuit, the network comprising a square millimeter of brain tissue, has been the subject of intense experimental and theoretical research. A full-scale model of the microcircuit at cellular and synaptic resolution [1] containing about 100,000 neurons and 300 million local synapses exhibits fundamental properties of in vivo activity.

Despite this success, the explanatory power of local models is limited, as interactions between areas contribute substantially to cortical dynamics. We therefore set out to construct a multi-scale spiking network model of all vision-related areas of macaque cortex that represents each area by a full-scale microcircuit with area-specific architecture. The layer- and population-resolved network connectivity integrates axonal tracing data from the CoCoMac database with recent quantitative tracing data, and is refined using dynamical constraints.

This research program raises methodological as well as technological questions: Are simulations at this scale feasible with available computer hardware [2]? Are full-scale simulations necessary, or can models of appropriately downscaled density be studied instead [3]? And finally: How can dynamical constraints be built into a high-dimensional spiking network model [4]?

The talk systematically addresses these questions and introduces the required technology before outlining the data integration process [5]. The simulation technology has been developed on the K computer in Kobe and JUQUEEN in Juelich and is incorporated in the NEST simulation code. Simulation results reveal a stable asynchronous irregular ground state with heterogeneous activity across areas, layers, and populations. Intrinsic time scales of spiking activity are increased in hierarchically higher areas, and functional connectivity shows a strong correspondence with that measured using fMRI. The model bridges the gap between local and large-scale accounts of cortex, and clarifies how the detailed connectivity of cortex shapes its dynamics on multiple scales.

[1] Potjans TC and Diesmann M (2014) The Cell-Type Specific Cortical Microcircuit: Relating Structure and Activity in a Full-Scale Spiking Network Model. Cerebral Cortex 24(3):785--806

[2] Kunkel S, Schmidt M, Eppler JM, Plesser HE, Masumoto G, Igarashi J, Ishii S, Fukai T, Morrison A, Diesmann M, Helias M (2014) Spiking network simulation code for petascale computers. Front. Neuroinform 8:78

[3] Van Albada S, Helias M, Diesmann M (2015) Scalability of Asynchronous Networks Is Limited by One-to-One Mapping between Effective Connectivity and Correlations PLoS Comput Biol 11(9): e1004490

[4] Schuecker J, Schmidt M, van Albada SJ, Diesmann M, Helias M (2017) Fundamental Activity Constrains Lead to Specific Interpretations of the Connectome PLoS Comput Biol 13(2): e1005179

[5] Schmidt, M., Bakker, R., Shen, K., Bezgin, G., Hilgetag, C.-C., Diesmann, M., van Albada, S.J. (2016) Full-density multi-scale account of structure and dynamics of macaque visual cortex. arXiv:1511.09364.

Bio: Prof. Dr. Markus Diesmann is Director of the Institute of Neuroscience and Medicine (INM-6, Computational and Systems Neuroscience), Director of the Institute for Advanced Simulation (IAS-6, Theoretical Neuroscience) and Director of the JARA-Institute Brain structure-function relationships (INM-10) at Jülich Research Centre, Germany. He is also full professor in Computational Neuroscience at Faculty of Medicine, RWTH University Aachen, Germany and affiliated with the Department of Physics of the same university.

Prof. Diesmann studied Physics at Ruhr University Bochum with a year of Cognitive Science at University of Sussex, UK. He carried out his PhD studies at Weizmann Institute of Science, Rehovot, Israel, and Albert-Ludwigs-University Freiburg, Germany. In 2002 he received his Ph.D. degree from Faculty of Physics, Ruhr University Bochum, Germany. From 1999 Prof. Markus Diesmann worked as senior staff at Dept. of Nonlinear Dynamics, Max-Planck-Institute for Dynamics and Self-Organization, Göttingen, Germany. He was an Assistant Professor of Computational Neurophysics at Albert-Ludwigs-University, Freiburg, Germany before gaining experience as a Unit Leader and Team Leader at RIKEN Brain Science Institute, Wako City, Japan. Since 2011 Prof. Markus Diesmann has been Director at INM-6 where he heads the Computational Neurophysics group. His main scientific interests include correlation structure of neuronal networks, models of cortical networks, simulation technology and supercomputing.


14:45–15:30 Jorge Mejias  (Center for Neural Science , New York University, USA)
Large-scale models of cortical dynamics: neural communication and cognitive computations

Abstract: In this talk, I will present a new set of computational approaches to model the brain of primates at a large-scale level of detail. We have based our models on a recently obtained connectivity data set of the macaque inter-areal connectivity, which provides information on the strength and directionality of inter-areal projections, and also on the position of each area in the cortical hierarchy. In particular, I will present (i) a model of a biophysical mechanism to explain the efficient propagation of neural signals across the cortical network, based on a balanced amplification of transient responses by cortical circuits, and (ii) a large-scale model to explain the emergence of working memory activity across different cortical areas and lobes, which provides a framework to understand distributed representations in the brain.

Bio: Jorge F. Mejias obtained his degree in Physics and Masters in Biomathematics from the University of Granada, Spain. In 2009, he obtained a PhD in Computational Neuroscience from the same university, under the supervision of Prof. Joaquin Torres and in close collaboration with Profs. Joaquin Marro, Hilbert Kappen and Nicolas Brunel. In 2010, Dr. Mejias moved to the University of Ottawa in Canada, to work as a postdoctoral fellow with Profs. Andre Longtin and Leonard Maler. In 2013, he moved to a postdoctoral position at the lab of Prof. Xiao-Jing Wang at New York University, where he is now an assistant research scientist. His main research interests include the dynamics of spiking neural networks and computational models of large-scale brain networks.


16:00-16:45 Subutai Ahmad (VP Research Numenta, USA) 
Why the Neocortex Has Layers and Columns, a Theory of Learning 3D Models of the World

Abstract: Neocortical regions are organized into columns and layers.  Connections between layers run mostly perpendicular to the surface suggesting a columnar organization, while some layers have long-range lateral connections that span columns. Similar patterns of connectivity exist in all regions but their exact role remains a mystery.  In this talk, we propose a network model composed of columns and layers that performs robust object learning and recognition. Each column processes a subset of the sensory input space and integrates its changing inputs over time. Lateral connections across columns allow the network to rapidly infer objects based on the partial knowledge of adjacent columns. Because columns integrate input over time and space the network can learn models of complex objects that extend well beyond the classic receptive field of individual cells. The model explains a number of empirical observations that have eluded theoretical understanding, including inter-layer and intra-layer connectivity patterns within cortical columns, and why some layers of cells have extensive lateral connections. Given the ubiquity of columnar and laminar connectivity patterns throughout the neocortex, we propose that all columns have more powerful recognition capabilities than previously assumed.

Bio: Subutai Ahmad is the VP of Research at Numenta, a research institution focused on understanding the computational principles of the neocortex, and developing technology for Machine Intelligence based on those principles.  Subutai’s research experience has focused on computational neuroscience, machine learning, and computer vision. Subutai holds a Bachelor's degree in Computer Science from Cornell University, and a Ph.D in Computer Science from the University of Illinois at Urbana-Champaign (thesis on computational neuroscience models of visual attention). 


16:45-17:30 Stefan Mihalas (Allen Institute for Brain Science, USA)
Cortical circuits implement optimal context integration and its gating

Neurons in the early cortex respond to a patch of the visual input, with modulation from the visual input from its surround [1]. This reflects the fact that features in natural scenes do not occur in isolation: lines, surfaces are generally continuous, and the surround provides context for the information in the classical receptive field. It is generally assumed that the information in the near surround is transmitted via lateral connections, between neurons in the same area [1]. A series of large scale efforts have recently described the relation between the lateral connectivity and visual evoked responses and found like-to-like connectivity between excitatory neurons [5, 6]. However current normative models of cortical function rely on sparsity [9], saliency [2] predict inhibition

between similarly tuned neurons. What computation are consistent with the observed structure of the lateral connections between the excitatory and diverse types of inhibitory neurons? 

We combined natural scene statistics [8] and mouse V1 neuron responses [3] to compute the lateral connections and computations of individual neurons which would optimally integrate information from the classical receptive field with that from the surround. The direct implementation requires single neurons to make complex computations on their inputs. While it is possible for such computations to be implemented by the dendritic trees, we show that an approximation can be achieved with relatively simple neurons. We show that this network has "like-to-like" lateral connections between excitatory neurons similar to the observed one [5, 6], distance dependence of connections similar to the observed ones [7]. Additionally, when these lateral connections are implemented in a neuronal network the reconstruction of natural scenes is significantly improved.

However sometimes we need to interpret scenes which have very different statistics. To preclude wrong priors from harming the representation, we include a system to gate this information. Overall, the system requires three classes of inhibitory neurons: one performing local normalization, one surround inhibition, and one gates the inhibition from the surround, similar to anatomical [4] and physiological studies [10].

We hypothesize that these computations: optimal integration of contextual cues and the capacity to gate them when uninformative are general property of cortical circuits, and the rules constructed for mouse V1 generalize to other areas and species.

[1] Alessandra Angelucci and Paul C. Bressloff. Contribution of feedforward, lateral and feedback connections to the classical receptive field center and extra-classical receptive field surround of primate V1 neurons. In Progress in brain research, volume 154, pages 93–120. 2006.

[2] Ruben Coen-Cagli, Peter Dayan, and Odelia Schwartz. Cortical Surround Interactions and Perceptual Salience via Natural Scene Statistics. PLoS computational biology, 8(3):e1002405, 3 2012.

[3] S. Durand, R. Iyer, K. Mizuseki, S. De Vries, S. Mihalas, and R.C. Reid. A comparison of visual response properties in the lateral geniculate nucleus and primary visual cortex of awake and anesthetized mice. Journal of Neuroscience, 36(48), 2016.

[4] X. Jiang, S. Shen, C. R. Cadwell, P. Berens, F. Sinz, A. S. Ecker, S. Patel, and A. S. Tolias. Principles of connectivity among morphologically defined cell types in adult neocortex. Science, 350(6264):aac9462–aac9462, 11 2015.

[5] Ho Ko, Sonja B Hofer, Bruno Pichler, Katherine A Buchanan, P Jesper Sjöström, and Thomas D Mrsic-Flogel. Functional specificity of local synaptic connections in neocortical networks. Nature, 473(7345):87–91, 5 2011.

[6] Wei-Chung Allen Lee, Vincent Bonin, Michael Reed, Brett J Graham, Greg Hood, Katie Glattfelder, and R Clay Reid. Anatomy and function of an excitatory network in the visual cortex. Nature, 532(7599):370–4, 4 2016.

[7] Robert B. Levy and Alex D. Reyes. Spatial Profile of Excitatory and Inhibitory Synaptic Connectivity in Mouse Primary Auditory Cortex. Journal of Neuroscience, 32(16), 2012.

[8] David Martin, Charless Fowlkes, Doron Tal, and Jitendra Malik. A Database of Human Segmented Natural Images and its Application to Evaluating Segmentation Algorithms and Measuring Ecological Statistics.

[9] B A Olshausen and D J Field. Emergence of simple-cell receptive field properties by learning a sparse code for natural images. Nature, 381(6583):607–9, 6 1996.

[10] Carsten K Pfeffer, Mingshan Xue, Miao He, Z Josh Huang, and Massimo Scanziani. Inhibition of inhibition in visual cortex: the logic of connections between molecularly distinct interneurons. Nature neuroscience, 16(8):1068–76, 8 2013.

Bio: Stefan Mihalas joined the Allen Institute as an Assistant Investigator in 2011 from Johns Hopkins University, where he was a postdoctoral fellow in neuroscience and subsequently an associate research scientist. As a computational neuroscientist, Mihalas has worked on models of both molecular and systems neuroscience including nervous system development, synaptic plasticity, minimalistic spiking neuron models, self-organized criticality, visual attention and figure ground segregation. Mihalas received his Diploma in physics and M.S. in mathematics from West University of Timisoara in Romania. He received his Ph.D. in physics from the California Institute of Technology.

Research: The central theme of my work is construction of models of how different neuronal circuits compute, linking structure to activity and computations. These ranged from models of molecular machinery aiming to explain the interactions between multiple spike timing dependent plasticity, models of the molecular machinery which allows very rapid development of the sympathetic nervous system and models of activity single neurons, to models of neuronal mechanisms underlying perceptual organization in the visual system and models of the influence of time perception in decision making. The focus of my current work is on describing the computational repertoire of the cortical microcircuits, and how such computations can be put together to describe the function of the cortical visual system in the mouse.

< Back