Brain Activity across Scales and Species: Analysis of Experiments and Simulations (BASSES)

EBRAINS Workshop

13–15 June 2022 | Rome & virtual

Download Preliminary Programme 2.5 MB

 

© Open Source & Roma Eventi

 
FAQ's and all you need to know
 

Registration is free of charge but mandatory . Attendance is open for interested scientists. Please make sure to register with your institutional email address.

To apply for a travel grant, please contact us at workshop.edu@humanbrainproject.eu no later than 6 May 2022. Students that submit an abstract will be prioritised in the evaluation process. 

The workshop will start on Monday 13 June 2022 at 13:00 CET and will end in the afternoon of Wednesday 15 June 2022. The programme comprises plenary sessions, hands-on sessions and social events. Please see the preliminary programme. 

The link to our virtual workshops platform as well as the online credentials will be sent out to all registered participants a few days prior to the workshop.

Yes, the virtual platform is accessible to all registered participants.

To make the virtual workshop an interactive experience, it is important that you have access to a stable internet connection, good audio (with microphone) and ideally (not mandatory) a webcam for video communication.

Also, to have the full workshop experience it is important that you have installed the Zoom Desktop Client application (v5.7.0 or higher) and that you access the workshop platform via the latest version of Google Chrome.

All times in the programme schedule are UTC+2/GMT+2= CEST. To prevent missed sessions, we recommend to use a time zone converter in advance.

If you have any questions about the event, please contact us at workshop.edu@humanbrainproject.eu.

The HBP and EBRAINS invite interested scientists to join the forthcoming EBRAINS Workshop on Brain Activities across Scales and Species.
The study of brain rhythms and of spatio-temporal patterns of brain activation is an important test-bench for understanding connectivity and the mechanisms that determine cognitive systems in mammals.
The large variety of available experimental protocols and measurement techniques enable researchers to investigate new scientific questions. In addition, the input from experimental observations is used to design theoretical models able to emulate brain dynamics and cognitive mechanisms in in-silico experiments.
The goal of the BASSES Workshop (Brain Activity across Scales and Species: analysis of Experiments and Simulations) is to provide an overview of the scientific topics of brain states and complexity, state transitions, and their connection with cognitive functions, and to demonstrate the achievements in this field obtained within the Human Brain Project thanks to the functionalities provided by the EBRAINS research platforms.
Lectures will showcase the latest advancements in analysis strategies and whole-brain modelling tools. Hands-On Sessions will allow the participants to be actively engaged and test the EBRAINS functionalities for data storage, curation, analysis, and modelling.
BASSES will allow people with different expertise, from experimental and theoretical neuroscientists to computer scientists, to share results and ideas and connect into a wider community.

Registration 

Registration deadline: 27 May 2022
Registration is free but mandatory.


Registration for on-site participation – free of charge but mandatory
  • Admission to all sessions
  • Workshop material
  • Coffee & lunch break during the workshop

The registration does not include travel, accommodation and participation in the social dinner. 
 

Registration for virtual participation - free of charge but mandatory
  • Admission to all streamed sessions
  • Workshop material

Confirmation of registration

After the registration has been completed you will receive an automatically generated notification via email.

Modification of the programme

The workshop organisers reserve the right to modify the programme.

Cancellation of the workshop

In the event that the workshop cannot be held or is postponed due to events beyond the control of the conference organisers (force majeure) or due to events that are not attributable to wrongful intent or gross negligence of the conference organisers, the workshop organisers cannot be held liable by delegates for any damages, costs, or losses incurred. Therefore, transportation costs, accommodation costs, costs for additional orders, financial losses, etc. will not be refunded.

Covid-19 information for on-site participation

  • Availability of seats for on-site participation is limited due to Covid-19 restrictions. In case the number of seats need to be reduced, participants will be invited on a first-come first-served basis. In such case you will be contacted by email.
     
  • Italy like other European countries has adjusted their COVID measurements to ease their restrictions. We still recommend to check entry requirements and current mask recommendations before booking your travels.
     
  • To attend the workshop in person, it is no longer mandatory to show an identity document and your EU Digital Covid Certificate / Green Pass, but FFP2 masks are required for indoor events. 
     
  • If the event due to changing Covid-19 restrictions has to be held fully virtual, participants who registered for on-site participation will be informed and automatically registered for the virtual workshop.
Logistics

Travel information

1. Rome by plane
Rome has two international airports that connect the city to many international destinations. The flight schedules can be found here.

2. Rome by train
In case you will travel to Rome by train, Roma Termini is the main train station. Detailed information can be found on the official website of the Italian railway.

3. Rome by coach
The bus companies Flixbus and Eurolines operate throughout Europe. Please find further information on their websites.

4. Rome by car
Rome is located along many major traffic routes, but driving into the city center of Rome can be quite an adventure and it’s not advisable. Therefore, public transport is recommended. Please note that the information provided on this site has been obtained from several different sources and therefore the organisers cannot accept any responsibility for errors therein.

Venue information
ROMA EVENTI – Fontana di Trevi Piazza della Pilotta
00187 Rome Italy

Accommodation
Please find a selection of nearby hotels here: Hotel suggestions by venue
Further hotels and other accommodations in Rome: https://www.turismoroma.it/en/node/18616

 

Practical information

Language
The official language in Rome and the rest of Italy is Italian.

Currency, credit cards, banking
The official currency in Italy is the EURO (EUR, €). The most common international credit cards (such as Mastercard or Visa) as well as debit cards (Maestro) are generally accepted. ATMs are available throughout the city.

Climate
The climate in Rome is Mediterranean, it can get very hot in summers.

Time zone
Italy is located in the Central European Time Zone (CET = GMT+1).

Please note that the information provided on this site has been obtained from several different sources and therefore the organisers cannot accept any responsibility for errors therein.

 

 

CALL FOR SUBMISSIONS

BASSES will have renowned scientists and early-career researchers side-by-side, thus enabling for the latters the opportunity to showcase their work and results. We invite original, high-quality submissions describing innovative research addressing the topics of brain dynamics, brain states, state transitions, complexity, cognitive functions, software solutions and analysis tools. The contributions can emphasize theoretical or empirical aspects. We particularly encourage submissions that leverage the functionalities provided by the EBRAINS research platforms or with a potential towards the integration with EBRAINS.
Poster sessions at the BASSES workshop will take place in a hybrid format and will be available for all event attendees to access digitally. More information will be sent to applicants upon acceptance of their poster abstract submission. A few abstracts will be selected to be presented in the plenary sessions.

Travel grant application is closed.

Extended abstract submission deadline: 20 May 2022

Create your own user feedback survey

If you don't see the submission form, please refresh your page or clicke here: https://www.surveymonkey.com/r/82VMLGQ

 

Special Issue "New Insights into Computational Neuroscience"

Apart from submissions to the workshop, participants are invited to submit their manuscripts to a related special issue of Applied Sciences “New Insights into Computational Neuroscience”, devoted to the illustration of new research in the field of Computational Neuroscience, focusing on its multidisciplinary aspects that gather expertise in physics and mathematics as well as biology, chemistry, engineering, and computer science. Submissions can be original research articles, review articles as well as short communications. 

Further details can be accessed on the Journal website: https://www.mdpi.com/journal/applsci/special_issues/Computational_Neuroscience1

Deadline for manuscript submissions: 20 September 2022.

Workshop participants are eligible to a discounted Article Processing Charge (APC) for publication in this open access journal. 

For further information please contact Dr. Giulia De Bonis

 

PRELIMINARY PROGRAMME

This programme is subject to change. Times displayed are in CEST/GMT+2/UTC+2.

Download Preliminary Programme 2.5 MB

12:30–14:00
Registration
 
14:00–16:30


Plenary Session I: Brain states and complexity
(chairs: Anna Letizia Allegra Mascaro & Giulia de Bonis)

When Causality meets Inference: complexity in neuroscience | Viktor Jirsa (Aix-Marseille University)

Models for bridging scales from neural circuits to the whole brain | Alain Destexhe (CNRS)

Student flash talks

Cortical Slow Waves: mechanisms, dynamics and modulation | Mavi Sanchez-Vives (IDIBAPS)

Title tbc | Marcello Massimini (University of Milan)
 

16:30–17:00
Coffee break
 
17:00–18:50


Plenary Session II: State transitions and their cognitive role
(Chair: Giulia de Bonis)

Cognitive and energetic benefits of awake/sleep cycles during incremental learning in multi-areal spiking neural networks
Pier Stanislao Paolucci (INFN)

A simple account of the complexity of slow wave activity | Maurizio Mattia (ISS)

Student flash talks

A comprehensive neural simulation of slow-wave sleep and highly responsive wakefulness dynamics | Jennifer Goldman (CNRS)

18:50–20:00
Informal discussions
 

 

9:00–10:45


Plenary Session III: Multi-scale approaches to investigate the brain complexity (I)
(Chair: Mavi Sanchez-Vives)

Title tbc | Johan Storm (University of Oslo)

Title tbc | Francesco Resta (LENS)

Loss of differentiation and complexity in the sleeping human brain: a multi-scale analysis | Andrea Pigorini (University of Milan)

Multiscale dynamical characterization of cortical brain states: integrating experimental and computational research in EBRAINS
Arnau Manasanch Berengué (IDIBAPS)

10:45–11:15
Coffee break
 
11:15–12:40

 

Plenary Session IV: Introduction to EBRAINS resources
(Chair: Pier Paolucci)

The transformative impact of the EBRAINS research infrastructure on research on brain structure, activity and cognitive function: Why - What - How 
| Jan Bjaalie (University of Oslo)

Title tbc | Michele Migliore

Student flash talks

12:40–13:40


Lunch break
 

13:40–15:20


Plenary Session V: Multi-scale approaches to investigate the brain complexity II.
(data analysis methods and results)

(Chair: Arnau Manasanch Berengué)

State-dependent cortex-wide broadcasting of sensory information | Elena Montagni (LENS)

Title tbc | Eric Landsness (University of Washington)

Blocks instead of puzzles pieces - analyzing cortical wave activity across scales in an adaptable framework
Robin Gutzen (Forschungszentrum Jülich)

Student flash talks

15:20–15:50
Coffee break
 
15:50–17:20


Hands-on Session I: Handling EBRAINS data

Lyuba Zehl | Forschungszentrum Jülich

17:20–18:50


Hands-on Session II: Running analysis in EBRAINS

Michael Denker, Robin Gutzen, Moritz Kern | Forschungszentrum Jülich

 

9:00–11:15


Plenary Session VI: Models and Simulation: Mean Field Simulation
(Chair: Jennifer Goldman)

Computational approaches to study cortical dynamics at multiple scales | Alessandra Camassa (IDIBAPS)

Cortical Slow Waves in Inferred Models of the Whole Hemisphere of Mouse | Chiara de Luca (INFN)

A general theory of cortical columns from first principle: out-of-equilibrium dynamics | Gianni Valerio Vinci (ISS)

Burst-dependent plasticity and dendritic amplification support target-based learning and hierarchical imitation learning | Cristiano Capone (INFN)

Title tbc | Thierry Nieus (University of Milan)

11:15–11:45
Coffee break
 
11:45–13:30


Plenary Session VII: Models and Simulation: Spiking Simulations
(Chair: Cristiano Capone)

High resolution wide-field spiking simulations of mouse cortical hemisphere | Elena Pastorelli (INFN)

Student Flash Talks

Multi-area full-density spiking network models of monkey and human cortices: from anatomy to resting-state dynamics
Sacha van Albada (Forschungszentrum Jülich)

Simulation of large-scale spiking network models on GPU systems: recent advances | Bruno Golosio (University of Cagliari)

Student Flash Talks

13:30–14:30


Lunch break

14:30–15:30


Hands-on Session III: Simulating spatially organised networks with NEST

Johanna Senk & Jasper Albers | Forschungszentrum Jülich

15:30–16:30


Hands-on Session IV: Validating models against data in EBRAINS

Andrew Davison & Shailesh Appukuttan | CNRS

16:30–17:00
Closing remarks
 

 

 

HANDS-ON SESSION DETAILS

Abstract
The session will focus on how to search, find and access data shared via EBRAINS. In a series of live-demonstrations the participants will learn how to interactively explore data registered in the EBRAINS Knowledge Graph database via the Knowledge Graph Search engine and the EBRAINS interactive atlas viewer „siibra-explorer“, how to learn and navigate the openMINDS metadata architecture that structures the information shared through the Knowledge Graph, as well as how to programmtically query and access data directly in the EBRAINS Knowledge Graph or based on their anatomical location through the EBRAINS atlases using siibra-Python. A proper documentation of all live-demonstrations will be handed out so that the participants can test and redo them outside of the session.

Workshop Chair & Speaker
Lyuba Zehl | Institute for Neuroscience and Medicine (INM-1), Jülich Research Centre, Jülich, Germany

Lyuba Zehl studied biology (BSc; topic: insect locomotion; supervisor: Proj. A. Büschges) and neuroscience (MSc; topic: whale brain anatomy; supervisor: Prof. W. Walkowiak) at the University of Cologne. In her doctoral studies at the RWTH Aachen University she studied systems neuroscience and (meta)data management at the Institute for Neuroscience & Medicine (INM-6) of the Jülich Research Centre (supervisor: Prof. S. Grün). As part of her thesis, she examined the spatio-temporal organisation of neural signals in the monkey motor cortex and refined the (meta)data management processes needed to collaborate on such complex neuroscience experiments. In 2017, she started as postdoctoral researcher in the Big Data Analytics Group of Prof. T. Dickscheid at the INM-1 of the Jülich Research Centre, working for the EBRAINS Data and Knowledge service. She is the product owner of the open Metadata Initiative for Neuroscience Data Structures (openMINDS) and, since 2020, co-leads the EBRAINS curation team.

 

Target audience
neuroscientists (basic programming knowledge is beneficial, but not necessary to follow the live-demonstrations)

Expected learning outcomes
Participants will learn how to interactively explore data via the EBRAINS Knowledge Graph Search engine (demo 1) and the EBRAINS interactive atlas viewer „siibra-explorer“ (demo 2), how to learn and navigate the metadata architecture of the Knowledge Graph (openMINDS; demo 3), how to programmatically query and access data directly in the EBRAINS Knowledge Graph (demo 4) or based on their anatomical location through the EBRAINS atlases (siibra-Python; demo 5).

Preparations
EBRAINS account required

Maximum number of participants
none

Abstract
In this EBRAINS hands-on session, we will explore how to represent and work with electrophysiological activity data containing spiking activity and local field potentials using the Neo data model, and how to perform a statistical analysis of such data using the Elephant framework. We will work with a fully annotated, realistic dataset that we obtain from the Knowledgegraph using the EBRAINS APIs and show a potential data analysis scenario step-by-step from loading the data to obtaining a final result using a series of prepared Jupyter notebooks. In a subsequent part of the hands-on session, we will explore how to implement and launch a more complex workflow on the EBRAINS RI by analyzing slow wave activity patterns using the modular COBRAWAP pipeline. Participants have the option to actively follow the live demonstrations, or recap the material later on their own.

Workshop Chair & Speakers
Michael Denker | Institute for Neuroscience and Medicine (INM-10), Jülich Research Centre, Jülich, Germany 
Robin Gutzen | Institute for Neuroscience and Medicine (INM-10), Jülich Research Centre, Jülich, Germany 
Moritz Kern | Institute for Neuroscience and Medicine (INM-10), Jülich Research Centre, Jülich, Germany 

Michael Denker received his diploma in physics from the University of Göttingen, Germany, in 2002. In 2004, he started as doctoral student in the Neuroinformatics and Theoretical Neuroscience lab of Sonja Grün at the Free University, Berlin. In 2006, Michael Denker relocated to Japan and became a researcher at the RIKEN Brain Science Institute, Wako-shi, Japan. He defended his PhD on relating coordinated spiking activity to local field potentials in 2009 at the Free University Berlin. In 2011 he joined the Institute of Neuroscience and Medicine (INM-6 and INM-10), Research Centre Jülich. In 2020, he became team leader for “Data Science for Electro- and Optophysiology Behavioural Neuroscience” at INM-10 to meet the upcoming challenges in the field of research data management in neuroscience at the national and international level.

 


 

Robin Gutzen studied physics at the RWTH in Aachen Germany where he concluded his M.Sc. with a thesis about validation frameworks for neural network simulation. Following his interest for computational neuroscience, since 2018, he is working as a PhD candidate with Sonja Grün and Michael Denker at the Research Center Jülich, where he currently investigates propagating cortical activity patterns and how to make data from heterogeneous sources comparable. He is further involved in the development of open-source software tools, and has a passion for data visualization and combining science and art.

 

 

 

 

Moritz Kern received the B.Eng. degree in biomedical engineering in 2019 and in 2021 the M.Sc. in applied computer science from the University of Applied Sciences, Ansbach. From 2019 - 2021 he was working as a Research Assistant with the Center for Signal Analysis of Complex Systems (CCS). His work included application and implementation of signal processing approaches in various scientific fields, especially time series analysis with respect to dynamic characteristics in biomedical applications. In 2021 he joined the Institute of Neuroscience and Medicine (INM-6), Research Centre Jülich and is currently working as Scientific Software Developer.

 

 

 

Target audience
neuroscientists (basic programming knowledge is beneficial, but not necessary to follow the live-demonstrations)

Expected learning outcomes
Participants will learn the basics of the Neo data model and API to represent neuronal activity data, including data loading, data annotation and data processing. Moreover, participants will learn the conceptual basis of performing rudimentary and advanced data analysis tasks using the Elephant toolbox. Particpants will learn how to utilize and design more complex analysis workflows using the EBRAINS Collaboratory.

Preparations
EBRAINS account required

Maximum number of participants
none

Abstract
This live demo ramps up from basic concepts to recent research on spatially organized neuronal network models with the simulator NEST (https://ebrains.eu/service/nest-simulator). Starting with the graphical user interface NEST Desktop (https://ebrains.eu/service/nest-desktop), we interactively construct a network model and explore its dynamics in the web browser. Afterwards, we turn to scripted PyNEST code and investigate the relationship between distance-dependent connectivity and spatially and temporally resolved patterns in spiking activity using Jupyter Notebooks. All examples will be made available in the EBRAINS Collaboratory.

Workshop Chair & Speaker
Johanna Senk | Institute of Neuroscience and Medicine (INM-6) and Institute for Advanced Simulation (IAS-6) and JARA-Institute Brain Structure-Function Relationships (INM-10), Jülich Research Centre, Jülich, Germany

Jasper Albers | Institute of Neuroscience and Medicine (INM-6) and Institute for Advanced Simulation (IAS-6) and JARA-Institute Brain Structure-Function Relationships (INM-10), Jülich Research Centre, Jülich, Germany

Johanna Senk is a post-doctoral researcher at the Institute of Neuroscience and Medicine (INM-6), Juelich Research Center, Germany. She graduated from RWTH Aachen University, Germany, with a Master's degree in physics in 2013, focusing on solid-state physics and computational physics. During her Master's studies, she spent an Erasmus exchange year at the University of Trieste, Italy. From 2014 to 2018, she performed her doctoral studies at INM-6 in the interdisciplinary field of computational neuroscience and received her PhD from RWTH Aachen University in 2019 (supervisor: Prof. Markus Diesmann). Her research is concerned with the modeling and simulation of large-scale spiking neuronal networks, fostering digitized and reproducible simulation-analysis workflows and addressing the performance of simulation technologies. She is interested in exploring the relationship between the spatial connectivity structure of the cortex and emerging dynamics.

 

 

Jasper Albers studied physics at RWTH Aachen University, spending an exchange semester at NTNU Trondheim in Norway. In his Master’s studies, he focused on quantum field theory and gauge theories and wrote his thesis on machine learning applications in cosmology. In December 2019 he joined INM-6 at Research Center Jülich where he is a PhD candidate under the supervision of Prof. Markus Diesmann and Prof. Sacha J. van Albada. His main research interests cover the fundamental principles of cerebral computation and the efficient encoding of information in the brain. In his work, Jasper focuses on modeling biologically inspired spiking neuronal networks at natural density, which includes gathering anatomical and physiological information and translating them into an abstract model that can be simulated. Specifically, he investigates the dynamical and functional implications of structured connectivity in a large-scale spiking neuronal network model of the macaque visual cortex.

 

 

Target audience
No prior knowledge is required. Basic programming skills are beneficial.

Expected learning outcomes
Simulating spatially organized networks with NEST Desktop and PyNEST.

Preparations
EBRAINS account required

Maximum number of participants
none

Abstract
This session will guide modellers and experimentalists in model validation (comparing simulation results to experimental data) using the EBRAINS Model Validation Framework.

Workshop Chair & Speaker
Andrew Davison | Paris-Saclay Institute of Neuroscience
Shailesh Appukuttan | Paris-Saclay Institute of Neuroscience

Dr Andrew Davison is a Senior Research Scientist and Neuroinformatics Group Leader in the Department of Integrative and Computational Neuroscience, Paris-Saclay Institute of Neuroscience (CNRS/Université Paris-Saclay, France). His main research interests are in large-scale, data-constrained, biologically-detailed modelling of neuronal networks.

 

 

 

 

Dr Shailesh Appukuttan is a postdoctoral researcher in the Department of Integrative and Computational Neuroscience, Paris-Saclay Institute of Neuroscience (CNRS/Université Paris-Saclay), France, with the Neuroinformatics research group. He is interested in the application of computational techniques to further biological research. As part of the Human Brain Project (HBP), his work involves the design and development of a model validation framework for neuroscience, and its integration into existing model development workflows, along with development of other related scientific tools. Personal webpage: http://www.shailesh-appukuttan.com/

 

 

Target audience
computational and experimental neuroscientists with at least a basic knowledge of Python programming

Expected learning outcomes
Participants will learn how to:

    • develop model-agnostic validation tests
    • adapt existing models so they can be more easily validated
    • use the EBRAINS model and test catalogue
    • register, search, view and compare the results of validation tests

Preparations
EBRAINS account required

Maximum number of participants
30

 

CONFIRMED SPEAKERS & SESSION CHAIRS
 

Jan Bjaalie, M.D., Ph.D., is Professor at the Institute of Basic Medical Sciences, University of Oslo, where he leads a team of researchers, data curation scientists, and software developers contributing to the building of the EBRAINS RI, the European distributed research infrastructure for brain and brain-inspired research. He is Infrastructure Director of the EU Human Brain Project, leader of the EBRAINS Data services, special advisor on neuroinformatics for the EBRAINS AISBL, Head of the Norwegian Neuroinformatics Node, and former Head of the Institute of Basic Medical Sciences at the University of Oslo (2009 - 2016). With a strong background in neuroanatomy and neuroscience, he is focused on making scientific research data more accessible and interpretable and on developing advanced brain atlasing tools for brain-wide analysis of multimodal data. In his role as founding Executive Director of the International Neuroinformatics Coordinating Facility (2006 – 2008), he initiated INCF programs on brain atlasing and multi-scale modeling. Professor Bjaalie has been partner and coordinator of several EU projects and has collaborated extensively with leading laboratories in many countries. He is Chief-Editor of Frontiers in Neuroinformatics and has served as member of the Neuroinformatics Committee of the Society for Neuroscience (2004 - 2009) and co-Chair (2018 - 2020) and Chair (2021) of the International Brain Initiative.

The transformative impact of the EBRAINS research infrastructure on research on brain structure, activity and cognitive function: Why - What - How

Abstract will follow.

Alessandra Camassa is a postdoctoral researcher at Sanchez-Vives lab which is part of the Systems Neuroscience group at IDIBAPS in Barcelona (Spain), where she recently defended her thesis under the joint supervision of Dr. M.V. Sanchez-Vives and Dr. M. Mattia. She studied biomedical engineering at Polytechnic University of Milan and during her master she spent a period as research trainee at the Harvard medical school where she joined the Division of Sleep and Circadian Disorders. During her PhD she studied the spatiotemporal dynamics of the cerebral cortex under unconscious brain states and during the transition towards consciousness at multiple scales, and she will be presenting some of the outcomes of her doctoral research. 

 

Computational approaches to study cortical dynamics at multiple scales

A. Camassa
Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Rosello 149-153, 08036 Barcelona, Spain
camassa@clinic.cat

An unsolved question of today’s neuroscience is to provide an understanding of brain dynamics under different brain states, while bridging the gap between multiple scales of observation in electrophysiological and modelling brain studies. Here, we present computational approaches suitable for the study of cortical dynamics in in vitro and in vivo electrophysiological neural recordings, as well as in simulated data from biophysically detailed spiking neural network models. We provide a comprehensive view through the study of network synchronization dynamics, waves propagation, temporal asymmetry, and dynamical richness of cortical activity under spontaneous and evoked conditions. Such multimethodology approaches allow us to integrate our results in a mechanistic framework, linking cortical properties emerging at the mesoscale with underling mechanisms observed at the microscale, in a fashion that highlights the value of the collaborative effort made by the researchers of HBP and EBRAINS consortia.

I am a postdoctoral researcher at the INFN in Rome. Our group is focused on theoretical and computational neuroscience and high-performance computing.
In the last years, I have been interested in understanding biological intelligence and in exploiting its most interesting aspects to improve artificial intelligence. I have proposed new biologically plausible algorithms for training recurrent spiking neural networks achieving unprecedented precision. Moreover, I demonstrated that the induction of "sleep" in an artificial NN reorganizes stored memories, improving after sleep performances.
I graduated from Sapienza University, on the topic: “inverse Ising methods applied to neural networks”. During my Ph.D. (at Sapienza University and the Italian Institute of Health), I developed mean-field models and large-scale spiking simulations of the cortical dynamics expressed under anesthesia, sleep, and during the transition to wakefulness.

 

Burst-dependent plasticity and dendritic amplification support target-based learning and hierarchical imitation learning

The brain can efficiently learn a wide range of tasks, motivating the search for biologically inspired learning rules. Most biological models are composed of point neurons and cannot achieve the state-of-art performances of artificial intelligence (e.g., they struggle to solve the credit assignment problem [1]).
Recent works have proposed that segregation of dendritic input (neurons receive sensory information and higher-order feedback in segregated compartments) [2] and generation of high-frequency bursts of spikes [1,4] would support backpropagation in biological neurons.
We propose that bursts and dendritic input segregation give the possibility to implement target-based learning [3] in a biologically plausible fashion and to orchestrate hierarchical imitation learning [5].

References

[1] Payeur, A, Guerguiev, J, Zenke, F et al. Burst-dependent synaptic plasticity can coordinate learning in hierarchical circuits. Nat Neurosci 24, 1010–1019 (2021).
[2] Guerguiev J, Lillicrap TP, Richards BA. Towards deep learning with segregated dendrites. eLife 2017;6:e22901
[3] DePasquale B, Cueva CJ, Rajan K, Abbott L, et al. full-FORCE: A target-based method for training recurrent networks. PloS one. 2018; 13(2):e0191527.
[4] Larkum ME. A cellular mechanism for cortical associations: an organizing principle for the cerebral cortex. Trends in Neurosciences. 2013; 36:141–151.
[5] Le, H, Jiang, N, Agarwal, A, Dudik, M, Yue, Y, and Daumé III, H. Hierarchical imitation and reinforcement learning. In International conference on machine learning (2018), pages 2917–2926. PMLR.

Chiara De Luca graduated in physics in 2019 at La Sapienza University and is currently perusing her PhD in behavioral neuroscience at La Sapienza University. 

 

Simulations Approaching Data: Cortical Slow Waves in Inferred Models of the Whole Hemisphere of Mouse 

Recent developments of new and powerful brain activity recording techniques, combined with the increasing anatomical knowledge provided by atlases and the growing understanding of neuromodulation principles, are allowing the study of  the brain at a whole new level, paving the way to the creation of data-driven extremely detailed effective network models.

In this work, aiming at reproducing the complex spatio-temporal dynamics of slow waves observed in experimental recordings of the cortical hemisphere of a mouse under ketamine anesthesia, a two-step inference procedure is proposed. First, inner loop, parameters of a mean-field model are optimized by likelihood maximization, exploiting anatomical knowledge to define connectivity priors; then, outer loop, the space of “external'' parameters is explored in search of an optimal match between the simulation outcome and the data, to enrich the spontaneous activity generated by the model with the desired variety of dynamical features observed in the data.  The validation of the model relies on the development of a versatile ensemble of analysis tools, applicable to both experimental and simulated data and capable of identifying and quantifying the spatio-temporal propagation of waves across the cortex; the tools enable the comparison of wave dynamics, evaluating the differences in the distribution of local observables (speed, direction, frequency), and, in the channel space, through the multivariate gaussian approach. 
Thanks to the interplay between data analysis and inference, the so-tuned model is capable of reproducing most of the non-stationary and non-linear dynamics displayed by biological networks.

Alain Destexhe is Professor and Research Director at CNRS, Director of the European Institute of Theoretical Neuroscience, and Vice-Director of the Paris-Saclay Institute of Neuroscience, one of the largest neuroscience institutes in France with 22 research teams. He is one of the initial writers of the Human Brain Project, and has been contributing theoretical models from the beginning of the project up to now. His lab studies brain states, from their genesis at the level of neuronal circuits, their organization at the level of large neuronal populations, and how they relate to sensory perception. He has published more than 160 journal articles, 2 monographs and 6 books as co-editor. He is also in the editorial board of several journals and co-Editor in Chief of the Journal of Computational Neuroscience.

Models for bridging scales from neural circuits to the whole brain

Modeling brain mechanisms is often confined to a given scale, such as single-cell models, network models or whole-brain models, and it is thus difficult to relate these models.  Here, we show an approach to build models across scales, starting from the level of circuits to the whole brain.  The key is the design of accurate population models derived from biophysical models of networks of excitatory and inhibitory neurons.  Such population models can be later integrated as units in large-scale networks defining entire brain areas or the whole brain.  We illustrate this approach by the simulation of asynchronous and slow-wave states, from circuits to the whole brain.  At the mesoscale (millimeters), these models account for travelling activity waves in cortex, and at the macroscale (centimeters), the models reproduce the synchrony of slow waves and their responsiveness to external stimuli.  This approach can also be used to evaluate the impact of sub-cellular parameters, such as receptor types or membrane conductances, on the emergent behavior at the whole-brain level.  This is illustrated with simulations of the effect of anesthetics.  We conclude by showing tools allowing users to simulate these paradigms in the EBRAINS platform.

Bio and picture to follow soon.

Jennifer Sarah Goldman is dedicated to understanding the human brain and engineering interventions to prevent cognitive dysfunction. She is fascinated by complex dynamics and aims toward an integrated understanding of human biology across space-time scales. Following the completion of her Bachelor of Arts studying Humanities at McGill University in Canada, Jennifer completed her PhD at the Montreal Neurological Institute and Hospital, using experimental methods to study cellular and molecular processes underlying brain development and degeneration. Jennifer completed her first postdoctoral appointment studying human neuroimaging data, including MRI and MEG in resting and active brain states with Alan Evans at the MNI. Finding information to connect knowledge from microscopic to macroscopic scales lacking, she began her second postdoctoral appointment, tasked with creating models of the human brain spanning molecules to whole brain dynamics, supervised by Professors Viktor Jirsa and Alain Destexhe. Jennifer’s work on multi-scale models has become one of the 5 showcases of the human brain project and she has won several awards for research and teaching during her career so far.

A comprehensive neural simulation of slow-wave sleep and highly responsive wakefulness dynamics

Hallmarks of neural dynamics during different human brain states span spatial scales from neuromodulators acting on microscopic ion channels to macroscopic changes in communication between brain regions. Developing a scale-integrated understanding of neural dynamics has therefore remained challenging. Here, we perform the integration across scales with mean-field modeling of Adaptive Exponential (AdEx) neurons, explicitly incorporating intrinsic, microscopic properties to endow excitatory and inhibitory neurons with membrane conductances describing membrane leaks, voltage-dependent synaptic communication, and spike-frequency adaptation. Here, we connect AdEx mean-field neural populations via structural tracts defined by the human connectome, and show that macroscopic, state-dependent dynamics resembling human brain activity emerge that mimic state-dependent signals during both spontaneous and evoked human brain activity. Specifically, modulating microscopic mechanisms enhancing spike-frequency adaptation leads to the emergence of slow-wave dynamics that synchronize across brain areas, shifting the functional connectivity, phase coupling, and power spectra, thus mimicking empirical data describing different brain states.  Moreover, evoked activity reproduces observations of more complex spread of stimuli between brain regions during wakefulness compared to slow-wave sleep, analyzed with the perturbational complexity index.  The model has been implemented with The Virtual Brain (TVB) simulator, and is open-access in EBRAINS, with publicly available tutorials that will be demonstrated during the talk. Our approach provides novel tools to evaluate how changes in microscopic parameters relate to large-scale emergent behavior in the brain, in particular enhanced responsiveness during awake compared to sleeping brain states, and thus offers a more unified, formal understanding of conscious and unconscious states and their associated pathologies. 

Bruno Golosio is associate professor of Applied Physics at the University of Cagliari and Research Fellow of the Italian Institute for Nuclear Physics. Since October 2017 he is director of the School of Medical Physics of the University of Cagliari. His main research interests are computational neuroscience, neural networks, computational methods for medical physics and biophysics, machine learning in medical imaging, Monte Carlo simulation methods. Over the last twenty years, he participated in several research projects, with increasing level of responsibility. Currently, he is principal investigator of the project icei-hbp-2020-0007 for MPI-GPU simulation of spiking neural network models in the Interactive Computing E-Infrastructure for the Human Brain Project (ICEI-HBP). He collaborates on the WaveScalES (wave scaling experiments and simulations) experiment of the Human Brain Project, and on the AIM (Artificial Intelligence in Medicine) project founded by the Italian Institute for Nuclear Physics (INFN). He is the first author of the software library NEST GPU for fast simulation of large-scale networks of spiking neurons (https://github.com/nest/nest-gpu) and of a cognitive neural model of language development, called ANNABELL (https://github.com/golosio/annabell/wiki). He also contributed to the development of the NEST simulator (http://www.nest-simulator.org). The trace of his research work is documented on about 120 research papers.

Simulation of large-scale spiking network models on GPU systems: recent advances

Spiking neuronal network models have proven to be a very powerful tool for studying the relationships between neuronal signals and high-level processes in the brain. Indeed, neuroscience research over the past several decades has shown that high-level brain functions are emergent properties of extremely large populations of neurons and that the dynamics of these populations exhibit a chaotic behavior, which cannot easily be reproduced in all its relevant features by simplified models. In recent years, multi-GPU systems and GPU clusters are establishing themselves as suitable systems to reconcile the need for fast simulations of very large-scale networks with that of having a high degree of flexibility in defining the model used to describe neuron dynamics, synaptic current and plasticity. In this talk I will present some of the most relevant recent developments in multi-GPU and GPU cluster simulations, focusing in particular on the problems of local and remote spike communication and scalability.

Robin Gutzen studied physics at the RWTH in Aachen Germany where he concluded his M.Sc. with a thesis about validation frameworks for neural network simulation. Following his interest for computational neuroscience, since 2018, he is working as a PhD candidate with Sonja Grün and Michael Denker at the Research Center Jülich, where he currently investigates propagating cortical activity patterns and how to make data from heterogeneous sources comparable. He is further involved in the development of open-source software tools, and has a passion for data visualization and combining science and art.

 

Blocks instead of puzzles pieces - analyzing cortical wave activity across scales in an adaptable framework
The expanding availability and variety of data and methodologies represent a great opportunity to access neural processes in finer detail. Leveraging the complementary insights from across experiments, species, and measurement techniques, however, poses a challenge as the data is too heterogeneous and the corresponding analyses too specific to allow for rigorous quantitative comparisons of the results. However, this challenge also promises new avenues of scientific progress. By aligning existing data and analyses from different sources in a reusable workflow we can build a broader basis for meta-studies, contextualization of individual studies, and model validation. Here, we showcase such an analysis pipeline with the application to cortical wave activity in the delta (‘slow waves’) and beta range, integrating capabilities to process diverse data and topical analytical methods within a consistent framework: the ‘collaborative brain wave analysis pipeline’ (Cobrawap). The pipeline design is based on modular building blocks that provide implementations of analysis methods and processing steps. The components are matched by their input-output relations and can be flexibly combined and arranged into a workflow to fit the requirements of the data and the scientific question. In this framework, by reusing the identical methods and implementations and by converging the heterogeneous data to a common descriptive level of wave activity, we are in a situation where analysis outcomes can be quantitatively compared using common characteristic measures. We demonstrate the versatility of the pipeline by analyzing slow wave activity in ECoG and calcium imaging recordings to evaluate the influence of dataset-specific parameters on the wave characteristics such as the type and dose of anesthesia or the measurement modality and their temporal and spatial resolution, and show that we can replicate corresponding findings from the literature. Furthermore, we show how the pipeline enables the benchmarking of methods by analyzing the same data with different method blocks and how the individual pipeline elements can be reused, rearranged, or extended to help derive analysis workflows for similar research endeavors and amplify collaborative research.

 

Viktor Jirsa is Director of the Inserm Institut de Neurosciences des Systèmes at Aix-Marseille-Université and Director of Research at the Centre National de la Recherche Scientifique (CNRS) in Marseille, France. Dr. Jirsa received his PhD in 1996 in Theoretical Physics and has since then contributed to the field of Computational Neuroscience with a focus on networks and dynamics. His research has been foundational for network science in medicine, leading to novel technologies for personalized brain modeling and clinical translation in epilepsy. Dr. Jirsa is lead of the brain simulation platform The Virtual Brain (www.thevirtualbrain.org) and a lead scientist in the European flagship Human Brain Project (https://www.humanbrainproject.eu/). Dr. Jirsa has been awarded several international and national awards including the Human Brain Project Innovation Prize (2021) and Grand Prix Départemental de Recherche en Provence (2018). Dr. Jirsa serves on various Editorial and Scientific Advisory Boards and has published more than 160 scientific articles and book chapters, as well as co-edited several books including the Handbook of Brain Connectivity.

 

When Causality meets Inference: complexity in neuroscience

Neuroscience is home to concepts and theories with roots in a variety of domains including information theory, dynamical systems theory, and cognitive psychology. Not all of those can be coherently linked, some concepts are incommensurable, and domain-specific language poses an obstacle to integration. Still, conceptual integration is a form of understanding that provides intuition and consolidation, without which progress remains unguided. This talk is concerned with the integration of deterministic and stochastic processes within an information theoretic framework, linking information entropy and free energy to mechanisms of emergent dynamics and self-organization in brain networks. We identify basic properties of neuronal populations leading to an equivariant matrix in a network, in which complex behaviors can naturally be represented through structured flows on manifolds establishing the internal model relevant to theories of brain function. The emergent perspective links concepts of symmetry and network theory and is illustrated along examples of healthy aging and epilepsy.

Arnau Manasanch is a Technical Coordinator in HBP WP2 and a Data Scientist and Data Manager at Sánchez-Vives Lab (IDIBAPS, Barcelona).
He studied Biomedical Engineering at UPF in Barcelona and obtained his M.Sc. at UB-UPC where he developed, as his thesis, an application to help clinicians precisely identify the origin of the epileptic source in DRE patients. His current research interests are in the analysis of electrophysiological signals obtained from in vivo, in vitro and in silico experiments in different animal models and under different brain states (conscious such as awake and unconscious such as asleep or anesthetized). His primary goal as HBP WP2 Technical Coordinator is to bridge the gap between the science developed in this Work Package, which comprises over 30 institutions from all over Europe, and the technologies developed in EBRAINS. Ultimately, his goal is to achieve an optimal integration within EBRAINS of the new scientific advances developed by the research groups, so that the neuroscience community at large can benefit from them.

 

Multiscale dynamical characterization of cortical brain states: integrating experimental and computational research in EBRAINS

The cerebral cortex spontaneously elicits different types of activity that changes over time according to the brain state. Brain states transitions from unconscious to conscious states are accompanied by changes in parameters such as cortical complexity, connectivity, synchronization and by a modulation of the excitatory-inhibitory network balance. The phenomenological correlates of such transitions have been observed in the cerebral cortex at multiple scales, i.e., at microscale in cortical slices in vitro, at mesoscale in cortical areas in vivo and at macroscale at whole brain level. The characterization of brain states transitions is fundamental to link the experimental and theoretical research outcomes to the clinic. A collaborative work is presented here as a review of the different studies that have well characterized the wide variety of multiscale brain states and their dynamics. This allows for an in-depth characterization of the spontaneous spatiotemporal dynamics of the cerebral cortex under different brain states at multiple scales through the study of waves propagation, network synchronization and complexity. The way in which this collaborative effort is integrated in the different EBRAINS services will also be presented. The content will be accessible and reproducible in EBRAINS, where different GUIs will allow the readers/users to 1) navigate and download the data and models, 2) reproduce the analysis and 3) visualize the results in an interactive way.

Maurizio Mattia: physicist by training with a PhD in Neurophysiology, he holds a permanent Researcher position at the Italian Institute of Health in Rome (Istituto Superiore di Sanità). He is also an adjunct professor of Neural Networks at the Physics Department of the “Sapienza” University of Rome. Since April 2016, he is core member of the FET Flagship “The Human Brain Project”. His interest is in bridging the gap between theory and experimental evidence on cortical network dynamics, by developing novel data analyses and theoretical approaches, with a focus on the collective dynamics underlying both the spontaneous activity under different brain states and the neuronal correlate of cognitive tasks like motor decision.

 

A simple account of the complexity of slow wave activity
Slow-wave activity (SWA) is a stereotyped activity pattern pervasively expressed during NREM sleep and deep anesthesia by the cerebral cortex of many animal species. As it spontaneously arises also in isolated in vitro (brain slices) and in vivo (cortical slabs) cortical networks, SWA is suggested to be the “default activity pattern” of the cerebral cortex: a collective phenomenon resulting from the interaction of its multiscale components. Indeed, SWA occurs in local cell assemblies as a slow oscillation between high-firing (Up) and almost quiescent (Down) states coordinated in space as waves of activity traveling across the whole cortical surface. The complexity of these ongoing wavefronts have been recently found to increase under anaesthesia as the awake state is approached. Here I will present recent findings on this topic from my lab eventually embedding them into a coherent theoretical framework. I will then talk about how sleep-wake transitions can be explained as a relatively simple modulation of some local features of cortical networks, eventually leading to predict instability patterns observed in experiment.

Dr. Elena Montagni obtained a master degree in Neurobiology in 2017 at the University of Pavia. From 2017 to 2020 she attended the International PhD in Atomic and Molecular  Photonics at University of Florence. During the PhD she investigated the functional organization of the motor cortex and its role in forelimb movement control by coupling optogenetics and calcium imaging to perform simultaneous all-optical manipulation and recording of mouse motor cortex. Since 2021 Dr. Montagni has held a post-doc fellowship at the European Laboratory for Non-Linear Spectroscopy (LENS) in Florence granted by the Human Brain Project (HBP). Her research interest is focused on slow wave propagation and how the cortical network integrates sensory information under different levels of anesthesia.

 

State-dependent cortex-wide broadcasting of sensory information

How the mouse cortex integrates sensory information over the large scale under anesthesia is still largely unexplored. To dissect how the brain state modulates the globally coherent waves of neuronal activity triggered by external stimuli, we performed wide-field optical imaging of evoked cortical dynamics in isoflurane anesthetized mice expressing GCaMP6f in excitatory neurons. We compared the cortical activity dynamics following peripheral sensory stimulation in two different brain states: deep and light anesthesia in the same chronically implanted mice. Our preliminary results show that the evoked response is composed by a primary stereotyped peak followed by a more time-distributed secondary response whose spatiotemporal features are influenced by the brain state.

Pier Stanislao PAOLUCCI serves as deputy leader of the “Networks underlying cognition and consciousness” WP in the Human Brain Project. Currently, his main interests are the modeling of the beneficial cognitive and energetic effects of sleep during incremental learning, the development of tools for the characterization and comparison of the spatio-temporal features of cortical waves expressed by actual brains and brain models and the study of the brain as inspiration for massive parallel computer design. Previously, he coordinated the European Projects EURETILE (2010-2015 on brain-inspired distributed computing) and SHAPES (2006-2009: co-design of Scalable Software Hardware Application Platforms). Between 1998 and 2009, he served as CTO of the design center of a leading semiconductor manufacturer, directing the design of the Diopsis family of Multi-Processor Systems-on-Chip. Since 1984, he is a member of the INFN APE parallel/distributed computing laboratory, where he contributed to the application-hardware-software co-design essential for several generations of APE parallel and distributed systems. He is the inventor and co-inventor of several international patents, numerical algorithms, and hardware-software co-design solutions. He holds a permanent research position at INFN (Istituto Nazionale di Fisica Nucleare) in Roma, Italy and he got his master’s degree in Physics at the Sapienza University of Roma.

 

Cognitive and energetic benefits of awake/sleep cycles during incremental learning in multi-areal spiking neural networks 

The alternation of wakefulness and sleep supports the brain energetic and cognitive efficiency in a large variety of high-level functions: among them, the capability of fast incremental learning from a few noisy examples, as well as the ability to associate similar memories in autonomously-created categories, the combination of contextual hints with sensory perceptions and the maintenance of the metabolic cost of brain functions within a budget notwithstanding the progressive increment in knowledge and performance. Sleep is known to be essential for awake performance, but the mechanisms underlying its role in supporting learning and energetic management are still to be clarified. This work leverages the recent experimentally driven hypotheses of apical isolation and apical drive principles to induce in a model some of the favourable energetic and cognitive effects associated to NREM and REM sleep, in compatibility with several experimental observation. Also, we follow the apical amplification concept to combine context and perception during training during awake learning. This way, we have recently added REM to the brain states accessible to our thalamo-cortical spiking models that already demonstrated the effects of incremental awake-NREM learning cycles. Specifically, we investigate both the effects of sleep on the internal synaptic structure of the network and on its neural activity in a two-area model. We demonstrate complementary homeostatic and associative effects of slow-wave and dreaming-like phases of sleep on cortico-cortical synapses and we show the consequent beneficial energetic consumption effects while keeping the sleep-induced cognitive effects.

Elena Pastorelli is a researcher at the National Institute of Nuclear Physics (INFN), where she is currently working on projects in the computational neuroscience area. Since 2016 she is involved in the Human Brain Project, with the aim to study Slow Wave Activity and the transition toward the awake state, using a multiscale approach that spans from the study of spatio-temporal activation patterns in a single cortical column, to the phenomenological analysis on the cortical surface. The specific focus of her activity is on Spiking Neural Network simulations of large-scale cortical models on HPC systems. 

 

High resolution wide-field spiking simulations of mouse cortical hemisphere 

Large-scale simulations of cortical models could help to shed light on the study of different brain states and cortical rhythms. In this talk, I will present high resolution wide-field spiking simulations of whole cortical hemisphere of a mouse at biological neural and synaptic density, reproducing the dynamics of Slow Wave Activity observed during anesthesia states. The model is obtained using parameters inferred from wide-field calcium imaging optical recordings acquired on anesthetized mice. Simulations are performed using two different spiking simulator engines (NEST and NEST GPU), and their outputs are analyzed using a tool suitable for the study of Slow Waves Activity. This methodology allows a direct comparison of outputs from different simulators, as well as between synthetic and experimentally recorded data, increasing the ability to refine and calibrate the model in terms of input parameters and biological plausibility. 

Dr. Andrea Pigorini received his PhD degree in Physiology from the University of Milan, Italy. He is currently a postdoctoral Research Fellow in Marcello Massimini’s Lab, at the Department of Clinical Sciences “L. Sacco”, University of Milan. His research focuses on the analysis of electroencephalographic activity recorded simultaneously at the scalp level and within the brain in treatment-resistant epileptic patients implanted with intracerebral electrodes for presurgical evaluation. During his doctorate he investigated the role of bistable dynamics in the reduction of the level of consciousness occurring during NREM sleep with respect to wakefulness. Before that, he studied biomedical engineering at the Politecnico di Milano (Italy) with an emphasis in advanced signal processing. His master degree thesis was focused on the development of a new method for time-frequency spectral analysis of EEG potentials evoked by transcranial magnetic stimulation.

 

Loss of differentiation and complexity in the sleeping human brain: a multi-scale analysis
INTRODUCTION.
Complexity, defined as the coexistence of functional differentiation and functional integration, is a general property of thalamo-cortical circuits that can be characterized at a multiscale level. Previous studies suggest that local dynamics occurring at the micro/meso-scale, such as sleep-like neuronal bistability, can be responsible for the collapse/emergence of global patterns of complex interactions among brain areas at the macro-scale. Here, we link the two scales by combining, for the first time, intracerebral single pulse electrical stimulation (SPES) with simultaneous invasive recordings (stereo-EEG, SEEG) and scalp high-density EEG (hdEEG) during both wakefulness and sleep. 

METHODS.
This work includes data collected during presurgical evaluation from 16 epileptic subjects. For each patient, SEEG was recorded through ~150 intracerebral contacts connected to a 192-channel amplifier. On top of this, simultaneous scalp recordings were obtained during the last day of hospitalization using a hdEEG (256 channels). Recordings of simultaneous SEEG and hd-EEG activity were combined with SPES during both wakefulness and Non-Rapid Eye Movement sleep (NREM). Specifically, a 5 mA current was applied through one pair of adjacent leads pertaining to the same depth electrode (i.e. one bipolar contact). A single stimulation session consisted of 30 consecutive biphasic pulses (positive-negative, 1/0.5ms) delivered at 0.5Hz. Data preprocessing was performed on both SEEG and hdEEG to exclude artefactual epochs and contacts from further analysis  and to remove stimulation artefacts.

RESULTS.
100 stimulation sessions were performed by delivering SPES during wakefulness at rest in different areas. 66 stimulation sessions were repeated during NREM sleep (stages N2 or N3; as confirmed by sleep scoring). At the group level, we show that: (1) The amplitude of the scalp-EEG response to SPES well correlates with the underpinning intracerebral activity (r2= 0.72 , p<0.05, mixed effects linear regression analysis). (2) The overall complexity, as assessed by the Perturbational Complexity Index was significantly higher in wakefulness with respect to NREM (p<0.05, mixed effects linear regression analysis). (3) The differentiation of the response to SPES, evaluated by performing the Principal Component Analysis across all sessions and comparing the ensuing number of components across states, was significantly higher in wakefulness with respect to NREM over the parietal areas (p<0.05; bootstrap statistics). (4) The reduction of complexity and differentiation was marked by the occurrence of an evoked slow wave (0.5-4Hz) associated with a suppression of high frequency power (>20Hz) that was significantly more prominent in NREM with respect to wakefulness (p<0.05, mixed effects linear regression analysis).

CONCLUSIONS.
We combined, for the first time, invasive and high-density non-invasive electrophysiological recordings in humans during intracerebral perturbations. This setup allows linking bistable dynamics (neuronal downstates in the Local Field Potential) and their effects on local cortico-cortical interactions with the overall connectivity and complexity assessed at the scalp level. Importantly, combining multiple stimulation sites across different states strongly suggests that during NREM sleep the reduction of complexity is mainly due to loss of differentiation.

Dr. Francesco Resta obtained a master’s degree in cellular and molecular biology in 2011 at the University of Florence. From 2012 to 2016 he attended the International Ph.D. in Pharmacology and Innovative Treatments granted by the Ministry of Education, Universities and Research at the University of Florence. Since November 2016 Dr. Resta holds a post-doc position at European Laboratory for Non-Linear Spectroscopy (LENS) granted by the Human Brain Project (HBP). His research interests focus on cortical connectivity and responsiveness in different brain states. In this context, his activities cover in vivo wide-field and two-photon imaging, electrophysiology and optogenetics.

Chair of Plenary Session IV

Mavi Sanchez-Vives, M.D., PhD in neurosciences, has been ICREA Research Professor at the Institute of Biomedical Research August Pi i Sunyer in Barcelona since 2008, where she leads the Systems Neuroscience group. After a PhD on biophysics of potassium channels, she moved on to work on the visual system (Rockefeller University, USA) and on thalamo-cortical circuits and dynamics (Yale University, USA). Her interests in the field of neuroscience are brain rhythms and brain states, their control by neuromodulation, and the investigation of novel brain interfaces. Her research emphasizes the integration of different levels -from ionic channels to brain networks- and a combination of an experimental and computational approach. She is author of over 200 articles in scientific journals and books. She has been involved in different international projects, including the Human Brain Project, in which she leads the work page devoted to “Networks underlying brain cognition and consciousness”.

 

Cortical Slow Waves: mechanisms, dynamics and modulation

Cortical slow waves are a multiscale phenomena that dominate global cortical dynamics in various conditions including slow wave sleep, deep anesthesia, some disorders of consciousness, or locally, in perilesional areas. This low complexity rhythmic pattern acts as an attractor that the system should escape in order to recover the levels of functional connectivity and complexity associated with conscious states. In this presentation we will discuss how the cortical network expresses these different dynamics, searching into cellular and network mechanisms that shape them. The combination of experiments and computational models will be helpful to disentangle such involved mechanisms considering that the cortex is a recurrently connected network. It will also be useful to analyze and understand the endogenous and exogenous neuromodulation (electrical, chemical, optical) that can influence the emergent activity, thus controlling the expressed state, excitability, complexity, processing power and communication across areas.

I’m a theoretical physicist who fell in love with neuroscience. The main question that I try to answer is always  “how does the brain work?” but under the lenses of Physics. In particular I’m fascinated by the dynamical aspects of the most complex system in the universe and the active role of noise in neural computations. Sometimes I dream of writing down the Newton laws of cortical columns.

 

A general theory of cortical columns from first principle: out-of-equilibrium dynamics
In this talk I will show recent developments in the theoretical description of cortical columns. The theory we derived is general and comes from first principles. Thus, it allows to bridge different scales relating, for example, single neuron details with neuronal population activity. Particular emphasis is given to the less studied out-of-equilibrium non-linear response of a cortical column to a signal that come either from other brain areas or from the external environment as in the case of TMS and TDSM simulations.

Sacha van Albada is Deputy Leader of HBP WP3, Junior Professor for Computational Neuroanatomy at the University of Cologne, and leader of the group "Theoretical Neuroanatomy" at Research Center Jülich, Germany. She obtained a Bachelor of Science degree from University College Utrecht, the Netherlands, followed by a Master's in Theoretical Physics from Utrecht University, and a PhD at the School of Physics at the University of Sydney, Australia. Her group combines anatomical and physiological data from a wide range of sources to build neural network models of mammalian cerebral cortex. The aim is to understand relationships between cortical structure and dynamics, and to provide models that serve as platforms for further refinement and for incorporating cortical function.

 

Multi-area full-density spiking network models of monkey and human cortices: from anatomy to resting-state dynamics

To approach a unified understanding of the structure and dynamics of the cerebral cortex, it is necessary to develop large-scale models at cellular resolution. This endeavor is increasingly enabled by existing simulation technology and large-scale anatomical datasets. In this talk, I will present large-scale spiking network models of macaque and human cerebral cortices. The models represent the full density of neurons and synapses in each local circuit. The simulations focus on the resting state as a first step toward cortical models bringing together a wide range of anatomical and physiological data and serving as test beds for hypotheses on cortical function. Finally, I will briefly present some recent work on connectivity concepts for neuronal network models, as well as theory and refinements of cortical models.

 

 
Jasper Albers | Forschungszentrum Jülich
Anna Letizia Allegra Mascaro | CNR & LENS
Shailesh Appukkutan | CNRS
Jan Bjaalie | University of Oslo
Alessandra Camassa | IDIBAPS
Cristiano Capone | INFN
Andrew Davison | CNRS
Giulia De Bonis | INFN
Chiara De Luca | INFN
Michael Denker | Forschungszentrum Jülich
Alain Destexhe | CNRS
Jennifer Goldman | CNRS
Bruno Golosio | University of Cagliari
Robin Gutzen | Forschungszentrum Jülich
Viktor Jirsa | Aix Marseille University
Moritz Kern | Forschungszentrum Jülich
Eric Landsness | University of Washington
Arnau Manasanch | IDIBAPS
Marcello Massimini | University of Milan
Maurizio Mattia | Italian National Institute of Health (Istituto Superiore di Sanità)
Michele Migliore | Institute of Biophysics, Italian National Research Council
Elena Montagni | LENS
Thierry Nieus | University of Milan
Pier Stanislao Paolucci | INFN
Elena Pastorelli | INFN
Andrea Pigorini | University of Milan
Francesco Resta | LENS
Mavi Sanchez-Vives | IDIBAPS
Johanna Senk | Forschungszentrum Jülich
Johan Storm | University of Oslo
Gianni Valerio Vinci | Italian National Institute of Health (Istituto Superiore di Sanità)
Sacha van Albada | Forschungszentrum Jülich
Lyuba Zehl | Forschungszentrum Jülich

 

Scientific Chairs & Local hosts
 

Dr. Anna Letizia Allegra Mascaro | CNR - National Research Council & LENS - European Laboratory for Non-Linear Spectroscopy, Firenze, Italy

Dr. Giulia De Bonis | INFN - National Institute of Nuclear Physics, Rome, Italy


Contact

workshop.edu@humanbrainproject.eu

 

Organised by

Logo-Education-Programme-Dick---black-on-white-161005_MG.jpg

 

In cooperation with

 

Venue

ROMA EVENTI – Fontana di Trevi
Piazza della Pilotta 4
00187 Rome
Italy