Week 1.  Network Neuroscience: from synapses to large-scale networks

Hippocampal-entorhinal circuits for spatial memory

Hauður Freyja Ólafsdóttir

The hippocampus is important for spatial and episodic memory. Place cells – the principal cell of the hippocampus – represent information about an animal’s spatial location. Yet, during sleep and rest, place cells spontaneously recapitulate (‘replay’) past trajectories. Replay has been hypothesised to serve a variety of functions in memory. In my talk, I will describe recent work I carried out, which showed replay may support a dual function: underpinning both spatial planning as well as the consolidation of new memories. Namely, we found during rest periods, place and grid cells from the deep medial entorhinal cortex (dMEC, the principal cortical output region of the hippocampus) replayed coherently. Importantly, putative dMEC replay lagged place cell replay by ~11ms; suggesting the replay coordination may reflect consolidation. Moreover, in a separate study, we found replay occurring just before movement to or upon arrival at a reward site preferentially depicted locations and trajectories consistent with the animals’ current task demands; perhaps indicative of spatial planning. However, we also found replay could dynamically ‘switch’ between a planning and consolidation mode in relation to engagement with task demands, and we found planning-like replay predicted the accuracy of imminent spatial decision. Finally, I will discuss unpublished work showing hippocampal-dMEC synchronisation in the theta-band may underlie hippocampal-dMEC replay coordination and ongoing work where we employ an ontogenetic approach to elucidating the neural circuit mechanisms of spatial memory.

—–

Dr. Hauður Freyja Ólafsdóttir received her PhD in Neuroscience from University College London (UK) in 2014 under the supervision of Prof. Hugo Spiers. Her thesis focused on the role of hippocampal place cell activity during sleep and wakefulness in spatial navigation. During her postdoc, in Dr. Caswell Barry’s (UCL) lab, she studied the contribution of hippocampal-cortical communication during different behavioral states to memory consolidation and spatial planning. In 2018 she received a Donders Mohrmann fellowship to start her own lab at the Donders Institute for Brain, Cognition and Behaviour (Radboud University, the Netherlands). The primary research objectives of her lab are to elucidate the neural circuit mechanisms underlying episodic and spatial memory, in development and adulthood, using state-of-the-art electrophysiological and imaging techniques.

Introduction to phase-amplitude coupling

Guido Nolte

This lecture covers mathematical concepts of coupling measures to estimate functional relations between neuronal oscillations partly within but mostly across different frequencies.

1. In the first part, I will give a very basic and largely qualitative introduction of fundamental measures to estimate the coupling of phases of oscillations at a specified frequency, with the most popular measures being the phase locking value (PLV) and coherence. I will also discuss the problem of volume conduction, meaning that functional relations can also be an artifact of signal mixing in sensor space, and how this can be addressed using e.g. the imaginary part of coherence.

2. In the main part of this lecture, I will discuss cross-frequency coupling and here basically phase-amplitude coupling. Several rather intuitive measures of phase-amplitude will be introduced. The main focus will then be on the relations between these coupling measures and bicoherence. Essentially a measure of coupling between phases at three different frequencies, which is a (normalized) third order statistical moment in the Fourier domain and which is analogous to coherence being a normalized second order statistical moment. It will be argued that the study of such relations is helpful to understand and optimize parameter setting for phase-amplitude coupling.

—–

Dr. Guido Nolte studied physics and made his PhD in 1995 at the University of Oldenburg/Germany. Since 1995 he has worked on methods development for the analysis of MEG and EEG data, in particular on forward and inverse calculations and on the construction of coupling measures. After working in Berlin, Albuquerque, Bethesda and Berlin again, he is now head of the MEG lab of the department of neurophysiology and pathophysiology at the UKE in Hamburg.

Models of large-scale brain dynamics

Joana Cabral

Brain activity exhibits chaotic signals comparable with the ones observed in networks of delay-coupled oscillators. On one side, transient brain rhythms are detected in EEG/MEG signals. On the other, fMRI reveals slow and spatiotemporally organized signal fluctuations. Using mathematical models of coupled oscillators, I will show how the network system can engage in a critical regime where intermittent cluster synchronization can generate signals sharing qualitative and quantitative features of human brain activity. I will present phenomenological insights obtained from models of coupled oscillators in the brain’s Connectome structure. Insights into general integrative processes of brain function will be discussed.

—–

Dr. Joana Cabral holds a PhD in Theoretical and Computational Neuroscience and a MSc+BSc in Biomedical Engineering, currently working as Assistant Researcher at the Life and Health Sciences Research Institute of the University of Minho in Portugal. She has visiting positions at the Center for Eudaimonia and Human Flourishing at the University of Oxford, UK, at the Shemesh lab at the Champalimaud Center for the Unknown in Lisbon, Portugal and at the Center for Music in the Brain at Aarhus University, Denmark. Her main interest is to investigate the fundamental principles underlying brain function. In other words, Joana is curious about the most primitive biophysical mechanisms at the genesis of coordinated brain activity, ultimately leading to our thoughts and actions. In her research, she tests hypothetical mechanistic scenarios using analytical and numerical methods to investigate fundamental brain mechanisms.

Brain simulations: from neural dynamics to cognitive functions

Jorge Mejias

The use of mathematical and computational models to study the brain has experienced a sharp rise in recent years as it constitutes a topic of interest, not only for neuroscience but for related disciplines like complex systems, nonlinear physics, psychology, and artificial intelligence. While many models have traditionally focused on ‘microscopic’ neural circuits (up to tens of thousands of neurons), novel data on detailed brain connectomes for human and no-human animals is facilitating the development of computational models of large-scale brain networks. In this lecture, I will cover recent work on large-scale brain modeling, and how it is used to understand the dynamics of neural activity observed in vivo and the perceptual and cognitive functions associated with such dynamics. In the first part, I will show how brain models are able to disprove predictions from local or microscopic computational models whose solutions do not hold then the large-scale recurrent connectivity of the brain is taken into account. A paradigmatical example is the propagation of sensory signals, a process traditionally studied with simplistic feedforward networks and which becomes dynamically unstable when real brain connectomes are considered. Proper solutions to this problem involve a strong balance between excitation and inhibition in brain networks. In the second part, I will show how this model can be extended to incorporate naturalistic neural oscillations, in an example of multiscale brain modeling. By constraining each level with neuroanatomical data and testing its dynamics with electrophysiological data, the multiscale model is able to reproduce a wide range of experimental observations across different spatial and temporal dynamics, and uncovering the emergence of functional hierarchical structures. Finally, in the third part, I will cover brain models which, besides realistic neural dynamics, can account for cognitive brain functions. In particular, I will introduce a data-constrained large-scale brain model able to explain the emergence of working memory –a particular type of short-term memory process in neural circuits –as a phenomenon distributed across large brain regions, in agreement with recent experimental evidence. This constitutes a new computational framework in which a large-scale model not only describes brain activity but also incorporates brain operations and function.

—–

Dr. Jorge Mejias is an assistant professor and head of the Computational Neuroscience Lab at the Cognitive and Systems Neuroscience Group at the University of Amsterdam in the Netherlands. With a background in physics and mathematics, he obtained a PhD in computational neuroscience from the University of Granada (Spain) in 2009. He then worked as a postdoctoral researcher at the University of Ottawa (Canada), New York University (USA), and the East China Normal University/ NYU Shanghai (China) before joining the University of Amsterdam in 2017. His research is focused on the theoretical and computational study of data-constrained multi-scale brain networks during perception and cognition, including brain functions such as working memory, multisensory integration, and predictive coding, as well as brain disorders that impair such functions. He also serves as an external member of the Institute Carlos I for Theoretical and Computational Physics in Granada, as a faculty member at the European Institute of Theoretical Neuroscience in Paris, and (currently) as director at the Organization for Computational Neurosciences (OCNS).

Excitation/inhibition balance as multi-scale mechanism regulating brain function

Klaus Linkenkaer-Hansen

The opposing forces of excitatory (E) and inhibitory (I) signaling fundamentally shape activity at many levels of neuronal organization. Heuristic arguments have favored a certain “E/I balance” to be important for normal brain function and “E/I imbalance” is thought to characterize many neurological and psychiatric disorders. The concept of E/I balance, however, is not uniquely defined at the mechanistic level because of many contributing factors, such as the size, number, and cellular distribution of synapses, the decay time of synaptic currents, and network topology. Here, we explain a definition of E/I balance that is inspired by computational modeling of critical brain dynamics. In this framework, E/I balance is an emergent network property—a functional state characterized by high spatio-temporal complexity emerging in neuronal networks balancing between order and disorder. The definition enables measuring an “E/I ratio” at the network level, and we show its utility for understanding basic principles of information processing in neuronal networks and for studying brain disorders in which E/I balance may be disrupted.

—–

Dr. Klaus Linkenkaer-Hansen is Associate Professor in the Department of Integrative Neurophysiology at Center for Neurogenomics and Cognitive Research (CNCR) at VU Amsterdam. He received his MSc in Physics-biophysics from the Niels Bohr Institute (Copenhagen University, Denmark) in 1998. He received his PhD from Helsinki University of Technology (Finland) in 2002. Following post-doctoral fellowships at the Netherlands Institute for Brain Research and CNCR, he became team leader of the Neuronal Oscillations and Cognition group in 2008. His research has been funded by 35+ awards from a number of European funding agencies, including the VENI scholarship from the Netherlands Organisation for Health Research and Development (NWO/ZonMW), the Physical Sciences (NWO/EW), the Social Sciences (NWO/MaGW), and the Technology Foundation (NWO/STW) of the Netherlands Organization for Scientific Research (NWO). His research is focused on the complex dynamics of neuronal oscillations and its implications for cognition in health and disease. Theories and methods from the physics of self-organization and complexity have guided his multi-disciplinary research since 1998. His research has shown that the complex temporal structure of ongoing oscillations is rich in information about the functional state and structure of the underlying neuronal networks.

Adaptive stimulation strategies

Hayriye Cagnan

Stimulation-based therapies for brain disorders provide an exciting avenue due to their focal and reversible nature. Current approaches face limits of side-effects, and effectiveness being restricted to a small subset of the patient population. During this talk, I will highlight recent work on developing and testing new therapies that aim to overcome these challenges.

—–

Dr. Hayriye Cagnan studied Electrical and Electronics engineering at Cornell University and specialized in signal processing and biomedical engineering (2000-2004). In 2004, she was accepted to the M.Sc. programme in Engineering and Physical Science in Medicine at Imperial College. Hayriye completed her Ph.D. in Neuroscience in a joint placement between the University of Amsterdam and Philips Research Laboratories in 2010. Subsequently, she joined the University of Oxford and worked on tremor pathophysiology. In 2015, Hayriye was awarded an MRC Skills Development Fellowship in Biomedical Informatics and worked on theoretical modelling of disease circuits. Hayriye was awarded an MRC Career Development Award in 2018 to establish her research group on Dynamic Neuromodulation.

Entering the neurorehabilitation market

Santiago Brandi

Going from the lab and controlled clinical trials to the real world poses additional challenges to neurorehabilitation solutions. In this talk, we will cover some of the main aspects neurorehabilitation products and companies must address for a successful commercial uptake.

—–

Eng. Santiago Brandi is an expert in technology for Neurorehabilitation and is the CEO of the spin-off company Eodyne.

Week 2. Brain Networks in Health and Disease: the Future of Neurotechnologies.

Past, presence, and future from neuroConn’s perspective.

The talk will focus on the history of neuroConn, the approach of neurocare for innovating mental health, and the future of having integrated solutions available for everyone.

Klaus Schellhorn

—–

Dr. Klaus Schellhorn has long-standing experience in neurostimulation & neurofeedback technology. He is the managing director of the company neuroConn GmbH that he co-founded as a spin-off of the Biomedical Engineering Department at the Technical University of Ilmenau in 2000.

Computational models to study the Parkinsonian brain and the mechanisms of neurostimulation techniques

Ciska Heida

Surgical treatment of neurological disorders like Parkinson’s disease, dystonia, and epilepsy was, until recently, mainly based on applying lesions at specific parts of the brain. While these procedures nowadays have been replaced by more reversible neurostimulation methods, most therapies for brain disorders are still based on trial-and-error, and effective mechanisms remain unknown. Using computational modeling can help us provide insight into neuronal network processes and interactions underlying normal and abnormal behaviour, as well as the mechanisms of therapeutic methods. While the difficulty in modeling is determining how much complexity needs to be included to simulate the aspects we are interested in as realistically as possible, it allows us to easily test new stimulation paradigms and stimulation targets. However, validation is essential for computational models to become useful tools for research purposes and clinical applications. In this presentation, two types of computational modeling will be presented and discussed in relation to experimental or clinical data: 1) 3D volume conduction modeling to study the local stimulation effects of deep brain stimulation (DBS) and motor cortex stimulation (MCS) in Parkinson’s disease; 2) neuronal network modeling, at a microscopic (cellular) and macroscopic (system) level, to study the behaviour of (part of) the basal ganglia-thalamocortical network under normal and Parkinsonian conditions and the effect of DBS.

—–

Dr. Ciska Heida is an Associate Professor in the Biomedical Signals & Systems group at the University of Twente, The Netherlands. She has a background in Electrical Engineering and received her PhD in the field of neurotechnology at the University of Twente. Her current research interests focus on increasing our understanding of the central mechanisms of human motor control, the pathophysiology underlying movement disorders, and the development and application of neuromodulation techniques for restoring motor control. Her research methods consist of computational modeling of (part of) the brain networks involved in motor control, and experimental research including movement tests performed by patients while recording brain activity. Her teaching activities are related to signal analysis of biomedical/clinical data and bioelectromagnetics.

Neurotechnology for Parkinson’s Disease

Richard van Wezel

Parkinson’s disease is one of the fastest-growing neurodegenerative diseases in the world. Advances in neurotechnology are necessary for a better understanding of the fundamental neurobiological mechanisms causing the disease, as well as for new approaches for diagnosis, and new devices for monitoring and treatment. In this seminar, I will provide an in-depth discussion of several of the new directions in this research field.

—–

Dr. Richard van Wezel is professor in Visual Neuroscience at the Radboud University Nijmegen and professor in Neurophysiology at Twente University (MedTech Center). He was director of the Donders Centre for Neuroscience in Nijmegen, and currently, he is vice-dean of Research at the Science faculty of Radboud University. Prof. van Wezel received many prestigious personal and team research grants (VIDI/NWO, High Potential UU, INTENS and other grants). He is co-chair of NeuroTech-NL (https://neurotech-nl.com/), a nation-wide consortium to promote Neurotechnology in the Netherlands. His main interest is in neurophysiology, visual perception, and brain plasticity and the applications of his fundamental research for clinical applications with a current applied focus on developing neurotechnology for Parkinson patients.

Feedforward and feedback processes that underlie visual perception

Matthew Self

Visual perception is the result of an interaction between the bottom-up information being conveyed via the retina and top-down processes reflecting our prior knowledge and expectations. In this lecture, I will explore how feedforward signals, either natural or artificially induced, can lead to conscious perception and how feedback signals organize our perception into the rich 3D percept we enjoy every day.

—–

Dr. Matthew Self is a senior researcher at the Netherlands Institute for Neuroscience in Amsterdam. After completing his studies at Cambridge University, he joined Professor Semir Zeki’s lab at University College London to study the integration of different visual features using fMRI. He then joined the lab of Pieter Roelfsema in Amsterdam to carry out post-doctoral work on the processing of feedback signals in the different layers of the visual cortex. His current work focuses on how feedback and feedforward processes interact in the visual cortex to produce visual perception.

The cerebral network of Parkinson’s disease tremor: a talk about effective connectivity

Michiel Dirkx 

Parkinson’s disease is characterized by the degeneration of dopaminergic cells in the midbrain, which leads to dopamine depletion in the striatum. One of the cardinal motor symptoms of Parkinson’s disease is a low-frequency (4-6 Hz) resting tremor. In the past, several studies have pointed toward the involvement of two cerebral circuits in the pathogenesis of Parkinson’s disease resting tremor: basal ganglia and a cerebello-thalamo-cortical circuit (consisting of cerebellum, thalamus and motor cortex). The exact circuit-level architecture of causal interactions between both circuits has long remained unclear, which has been a topic for our research over the last couple of years. In this presentation, we will talk about how we used functional MRI together with computational modelling (i.e. Dynamic Causal Modelling) to study the cerebral network of Parkinson’s disease resting tremor. We will focus on tremor-related cerebral activity and how different cerebral regions influence each other, i.e., how they are effectively connected. We propose the dimmer-switch hypothesis where the basal ganglia initiate a tremor episode, and the cerebello-thalamo-cortical circuit produces tremor amplitude. Next, we will discuss how this system-level framework has formed the basis to study several clinical characteristics of Parkinson’s disease tremor, such as the influence of dopaminergic medication and cognitive stress. Finally, we will discuss how this knowledge can be used to improve the therapy of Parkinson’s disease tremor.

—–

Dr. Michiel Dirkx is a post-doctoral researcher at the Donders Centre for Cognitive Neuroimaging and a neurologist in training at the Radboud University Medical Centre. He obtained his medical degree at the Radboud University Nijmegen in 2018, and his PhD thesis (“Neural mechanism of Parkinson’s tremor”) in 2020. Since 2020 he has worked as a post-doctoral researcher in the group of dr. Rick Helmich (“Systems Neurology”) where he focusses on the pathophysiology of movement disorders such as Parkinson’s disease. For this, he uses methods such functional MRI, electrophysiological tools such as EMG/accelerometry and computational modelling (Dynamic Causal Modelling).

Lessons for neurorobotics from bio-inspired learning and control principles of the central nervous system

Silvia Tolu

This talk will present some principles for the design of neurorobotics systems to reproduce animal behaviours and test brain theories. I will for example refer to the embodied intelligence principle that refers to the capability of living systems to learn, act, and adapt during interactions with dynamic environments. Neuro-robots become able to learn and adapt their tasks in complex and changing environments by means of bio-inspired control systems in which the brain is embedded. We will discuss how computational models in motor control and model learning architectures, including the internal model learning theory, can be applied to build bio-inspired control systems. Results represent a step closer towards a better understanding of the underlying neural mechanisms for motor control and learning, but also the starting point for achieving collaborative robots acting in complex applications such as manufacturing, assistive living, or as a tool for discovering the root causes of neurodegenerative diseases.

—–

Dr. Silvia Tolu is an Associate Professor and currently leads the Neurorobotics team at the Technical University of Denmark (DTU). She holds a Ph.D. degree in Neuromorphic Computing from the University of Granada (Spain). Her main research experience is in bio-inspired control for robots, compliant human-robot interactions, computational models of the cerebellum, and machine learning. Her research focuses on developing brain-based methods and technologies for enabling autonomy in robots while they act in realistic, dynamic, and rich sensory environments. Her vision is to open revolutionary robotics paradigms and to discover novel methods for diagnosis and rehabilitation of neuro-degenerative diseases. In 2016, she was granted with an IF-EF H2020 Marie Sklodowska-Curie Fellowship (BIOMODULAR, Project ID: 70510 2017-2019). She has been principal investigator in the Co-Design Project 2 (CDP2 Human Brain Project HBP-SGA2) about cerebellar implementations in robotic systems (2018-2020). Silvia has published in influential journals and proceedings in both robotics and computational neuroscience and has served as reviewer for top-ranked conferences and journals. She is currently in the Editorial Board of Frontiers Neurorobotics.

Multimodal Neuroimaging with CURRY

Fernando Gasca

This presentation will cover how the CURRY software integrates multiple imaging modalities (MRI, CT, PET, SPECT, fMRI, DTI) together with MEG/EEG data for offering various insightful possibilities when analyzing the brain. We will go through practical scenarios in which CURRY is useful for obtaining the most out of the available data: (a) reviewing multiple co-registered imaging modalities simultaneously, (b) creating anatomically accurate and subject-specific head models from MRI and DTI in order to perform source analysis, (c) pre-operative planning of stereo-EEG (intracranial) electrode implantation and post-operative evaluation, (d) combined MEG/EEG data and source analysis.

—–

Dr. Fernando Gasca graduated from the Iberoamerican University in Mexico City with a degree in Biomedical Engineering. He was co-founder of CODE Ingeniería, a Mexico City-based technology development company, where he worked as an engineer. He later received his Ph.D. in Neuroengineering from the University of Lübeck, Germany. His research focused on the modeling of transcranial stimulation techniques. Since 2014 he joined Compumedics Europe GmbH, Compumedics Neuroscan, as part of the CURRY development team in Hamburg, Germany. He has been involved in the development of the Image and Signal Processing modules, as well as the Finite Element Method (FEM) functionality and automatic spike and seizure detection. He is presently at 

Brain networks in stroke and cerebral small vessel disease

Bastian Cheng

I will discuss MRI connectome analysis and clinical implications in our prospective stroke cohorts and the Hamburg City Health Study (HCHS), the local, large-scale epidemiological study.

—–

Dr. Bastian Cheng is a consultant neurologist and associate professor at the University Medical Hospital in Hamburg. His clinical focus lies on the diagnosis and treatment of patients with stroke and cerebrovascular diseases. Scientific key interests are structural and functional magnetic resonance imaging in stroke, multimodal approaches of lesion-symptom inference as well as modelling and analysis of localized and large-scale structural brain networks in stroke and cerebrovascular disease relating to clinical recovery.