Full Professor, Bertarelli Foundation Chair in Translational Neuroengineering

EPFL

Switzerland

Title: Intraneural Neuroprostheses to Understand and Restore Sensory, Motor, and Autonomic Neural Functions

Abstract: Neuroengineering is a novel discipline combining engineering including micro and nanotechnology, electrical and mechanical, and computer science with cellular, molecular, cognitive neuroscience with two main goals: (i) increase our basic knowledge of how the nervous system works; (ii) develop systems able to restore functions in people affected by different types of neural disability. In the past years, several breakthroughs have been reached by neuroengineers in particular on the development of neurotechnologies able to restore sensorimotor functions in disabled people.
In this presentation, I will provide several examples on how implantable interfaces can be used to restore sensory (tactile feedback for hand prostheses, vision), motor (grasping, locomotion), and autonomic functions (for type 2 diabetes and cardiovascular problems) and how they can be used also to understand cognitive functions such as language and decision making.

Biography: Silvestro Micera is currently Professor of Bioelectronics at the Scuola Superiore Sant’Anna (SSSA, Pisa, Italy) and at the Ecole Polytechnique Federale de Lausanne (Lausanne, Switzerland) where he is holding the Bertarelli Foundation Chair in Translational NeuroEngineering. He received the University degree (Laurea) in Electrical Engineering from the University of Pisa, in 1996, and the Ph.D. degree in Biomedical Engineering from the Scuola Superiore Sant’Anna, in 2000. From 2000 to 2009, he has been an Assistant Professor of BioRobotics at the Scuola Superiore Sant’Anna. In 2007, he was a Visiting Scientist at the Massachusetts Institute of Technology, Cambridge, USA with a Fulbright Scholarship. From 2008 to 2011 he was the Head of the Neuroprosthesis Control group and Group Leader at the Institute for Automation, Swiss Federal Institute of Technology, Zurich, CH. He was the recipient of the “Early Career Achievement Award” and of the “Technical Achievement Award” of the IEEE Engineering in Medicine and Biology Society in 2009 and 2021, respectively. Dr. Micera’s research interests include the development of neuroprostheses based on the use of implantable neural interfaces with the central and peripheral nervous systems to restore sensory and motor function in disable persons. In particular, he is currently involved in translational experiments for hand prosthesis control in amputees, and the restoration of vestibular function, grasping and locomotion in different neurological disorders. He is author of more than 300 WoS peer-reviewed papers and several international patents. He is currently Associate Editor of IEEE Transactions on Neural Systems and Rehabilitation Engineering and of IEEE Transactions on Medical Robotics and Bionics. He is also member of the Editorial Boards of the Journal of Neuroengineering and Rehabilitation, and of Journal of Neural Engineering.

Institute Head, Institute of Theoretical Computer Science

Graz University of Technology

Austria

Title: Spiking Neural Networks for Neuromorphic Computing

Abstract: The success of deep learning has proven the power of neural networks as a trainable computational artefact. While artificial neural networks (ANNs) are biologically inspired, they differ in many aspects from their biological counterparts. One salient aspect is that biological neurons communicate with short voltage pulses, so-called spikes. A variety of models for such spiking neurons exist and the resulting networks are called Spiking Neural Networks (SNNs). SNNs have several potential advantages over ANNs: they are provably more computationally powerful and they can give rise to more energy-efficient implementations in specialised neuromophic hardware. However, SNNs were notoriously hard to train, rendering these potential advantages insignificant. In this talk, I will discuss recent progress on SNN training and propose several possibilities to further improve their computational capabilities. Finally, I will show that even simple SNNs can be an effective tool for bio-signal-processing.

Biography: Robert Legenstein received his PhD in computer science from Graz University of Technology, Graz, Austria, in 2002. In 2010 he got his Habilitation (venia docendi) for neuroinformatics. Currently he is full professor at the Department of Computer Science, TU Graz and head of the Institute for Theoretical Computer Science. His primary research interests are learning in models for biological networks of neurons and neuromorphic hardware, probabilistic neural computation, brain-inspired machine learning, and memristor-based computing concepts. Robert Legenstein serves as action editor for Transactions on Machine Learning Research and has served as associate editor of IEEE Transactions on Neural Networks and Learning Systems. He was several times area chair for ICLR and Advances in Neural Information Processing Systems (NeurIPS). Robert Legenstein is a board member of the Austrian Society for Artificial Intelligence.

Director of the Center of Brain and Cognition

Pompeu Fabra University

Spain

Title: The Thermodynamics of Mind

Abstract: Finding precise signatures of different brain states is a central, unsolved question in neuroscience. We reformulated the problem to quantify the ‘inside out’ balance of intrinsic and extrinsic brain dynamics in brain states. The difference in brain state can be described as differences in the detailed causal interactions found in the underlying intrinsic brain dynamics. We used a thermodynamics framework to quantify the breaking of the detailed balance captured by the level of asymmetry in temporal processing, i.e. the arrow of time. Specifically, the temporal asymmetry was computed by the time-shifted correlation matrices for the forward and reversed time series, reflecting the level of non-reversibility/non-equilibrium. We found precise, distinguishing signatures in terms of the reversibility and hierarchy of large-scale dynamics in three radically different brain states (awake, deep sleep and anaesthesia) in electrocorticography data from non-human primates. Significantly lower levels of reversibility were found in deep sleep and anaesthesia compared to wakefulness. Non-wakeful states also showed a flatter hierarchy, reflecting the diversity of the reversibility across the brain. Overall, this provides signatures of the breaking of detailed balance in different brain states, perhaps reflecting levels of conscious awareness.

Biography: Gustavo Deco is Research Professor at the Institució Catalana de Recerca i Estudis Avançats (ICREA) and Professor (Catedrático) at the Pompeu Fabra University (UPF) where he leads the Computational Neuroscience group. He is also Director of the Center of Brain and Cognition (UPF). In 1987 he received his PhD in Physics for his thesis on Relativistic Atomic Collisions. In 1987, he was a postdoc at the University of Bordeaux in France. From 1988 to 1990, he obtained a postdoc of the Alexander von Humboldt Foundation at the University of Giessen in Germany. From 1990 to 2003, he leads the Computational Neuroscience Group at Siemens Corporate Research Center in Munich, Germany. He obtained in 1997 his Habilitation (maximal academical degree in Germany) in Computer Science (Dr. rer. nat. habil.) at the Technical University of Munich for his thesis on Neural Learning. In 2001, he received his PhD in Psychology at the Ludwig-Maximilians-University of Munich.

Director of the Centre for Electronics Frontiers

University of Edinburgh

UK

Title: Bio-inspired Memory Technologies

Abstract: Every day globally, we generate about 2.5 quintillion bytes of data – equivalent to around five million household hard disks of storage. Our demand for electronic devices that can instantly store and process huge quantities of information are only set to increase with the advent of data-driven technologies such as the Internet of Things (IoT) and artificial intelligence. Such technologies, however, come with a growing environmental cost, due to their intensive use of energy and natural resources.
In stark contrast, our brains operate in a complex way to remember information, allowing us to repeatedly write new memories atop old ones in a “palimpsest” fashion – the name given to historical manuscripts on which later writing are superimposed on earlier writing. We have this extraordinary ability thanks to biochemical processes found in biological synapses, which allow us to learn quickly and continuously – at multiple timescales – without forgetting old memories. In his lecture, Prof Themis Prodromakis will present the attributes of memristive technologies that make this emerging technology attractive for a variety of applications with a focus on bio-inspired memories.

Biography: Themis holds the Regius Chair of Engineering at the University of Edinburgh and is Director of the Centre for Electronics Frontiers. His work focuses on developing metal-oxide Resistive Random-Access Memory technologies and related applications and is leading an interdisciplinary team comprising 30 researchers with expertise ranging from materials process development to electron devices and circuits and systems for embedded applications. He holds a Royal Academy of Engineering Chair in Emerging Technologies and a Royal Society Industry Fellowship. He is an Adjunct Professor at UTS Australia, visiting Professor at the Department of Microelectronics and Nanoelectronics at Tsinghua University, and Honorary Fellow at Imperial College London. He is Fellow of the Royal Society of Chemistry, the British Computer Society, the IET and the Institute of Physics and is also Senior Member of the IEEE. He served as the Director of the Lloyds Register Foundation International Consortium for Nanotechnology and Co-Director of the UKRI Centre for Doctoral Training in Machine Intelligence for Nano- Electronic Devices and Systems (MINDS). In 2015, he established ArC Instruments Ltd that delivers high-performance testing infrastructure for automating characterisation of novel nanodevices in over 21 countries and in 2019 he founded SoneT.ai that is building new power-efficient AI hardware solutions. His contributions in memristive technologies and applications have brought this emerging technology one step closer to the electronics industry for which he was recognised as a 2021 Blavatnik Award UK Honoree in Physical Sciences and Engineering.

Chair of Highly-Parallel VLSI-Systems and Neuromorphic Circuits

Technische Universität Dresden

Germany

Title: Bio-Inspired AI from Edge to Cloud: SpiNNaker 2 and Beyond

Abstract: AI is having an increasingly large impact on our daily lives. However, current AI hardware and algorithms are still only partially inspired by the major blueprint for AI, i.e. the human brain. In particular, even the best AI hardware is still far away from the 20W power consumption, the low latency and the unprecedented large scale, high-throughput processing offered by the human brain.
In this talk, I will describe our bio-inspired AI hardware and algorithms, from our award-winning edge systems up to SpiNNaker2, which is the largest cloud system for real-time AI worldwide.
On the hardware side, we combine high-performance and high-efficiency machine learning with sensor/actuator processing at millisecond latency. On the algorithm side, we merge the efficiency of biological approaches with the data handling capability of deep neural networks and the robustness of symbolic AI. Thus, our bio-inspired AI hardware&software framework represents a breakthrough in the area of real-time human-machine interaction at cloud scale, e.g. for automotive, smart city or tactile internet applications.

Biography: Christian Mayr is a Professor of Electrical Engineering at TU Dresden. From 2003 to 2013, he has been with Technische Universität Dresden, with a secondment to Infineon (2004-2006). From 2013 to 2015, he was a group leader/postdoc at the Institute of Neuroinformatics, University of Zurich and ETH Zurich, Switzerland. Since 2015, he is head of the Chair of Highly-Parallel VLSI-Systems and Neuromorphic Circuits at Technische Universität Dresden. His research interests include computational neuroscience, bio-inspired artificial intelligence, brain-machine interfaces, AD converters and general System-on-Chip design. He is author/co-author of over 100 publications and holds 4 patents. He is a PI in two German excellency clusters, in the national AI supercomputing center Scads. AI and in the EU Flagship Human Brain Project. He is one of the co-founders of SpiNNcloud Systems GmbH, which commercializes the largest real time AI machine worldwide, the SpiNNaker2 bio-inspired supercomputer.