-
Increasing Liquid State Machine Performance with Edge-of-Chaos Dynamics Organized by Astrocyte-modulated Plasticity
Authors:
Vladimir A. Ivanov,
Konstantinos P. Michmizos
Abstract:
The liquid state machine (LSM) combines low training complexity and biological plausibility, which has made it an attractive machine learning framework for edge and neuromorphic computing paradigms. Originally proposed as a model of brain computation, the LSM tunes its internal weights without backpropagation of gradients, which results in lower performance compared to multi-layer neural networks.…
▽ More
The liquid state machine (LSM) combines low training complexity and biological plausibility, which has made it an attractive machine learning framework for edge and neuromorphic computing paradigms. Originally proposed as a model of brain computation, the LSM tunes its internal weights without backpropagation of gradients, which results in lower performance compared to multi-layer neural networks. Recent findings in neuroscience suggest that astrocytes, a long-neglected non-neuronal brain cell, modulate synaptic plasticity and brain dynamics, tuning brain networks to the vicinity of the computationally optimal critical phase transition between order and chaos. Inspired by this disruptive understanding of how brain networks self-tune, we propose the neuron-astrocyte liquid state machine (NALSM) that addresses under-performance through self-organized near-critical dynamics. Similar to its biological counterpart, the astrocyte model integrates neuronal activity and provides global feedback to spike-timing-dependent plasticity (STDP), which self-organizes NALSM dynamics around a critical branching factor that is associated with the edge-of-chaos. We demonstrate that NALSM achieves state-of-the-art accuracy versus comparable LSM methods, without the need for data-specific hand-tuning. With a top accuracy of 97.61% on MNIST, 97.51% on N-MNIST, and 85.84% on Fashion-MNIST, NALSM achieved comparable performance to current fully-connected multi-layer spiking neural networks trained via backpropagation. Our findings suggest that the further development of brain-inspired machine learning methods has the potential to reach the performance of deep learning, with the added benefits of supporting robust and energy-efficient neuromorphic computing on the edge.
△ Less
Submitted 26 October, 2021;
originally announced November 2021.
-
BioGrad: Biologically Plausible Gradient-Based Learning for Spiking Neural Networks
Authors:
Guangzhi Tang,
Neelesh Kumar,
Ioannis Polykretis,
Konstantinos P. Michmizos
Abstract:
Spiking neural networks (SNN) are delivering energy-efficient, massively parallel, and low-latency solutions to AI problems, facilitated by the emerging neuromorphic chips. To harness these computational benefits, SNN need to be trained by learning algorithms that adhere to brain-inspired neuromorphic principles, namely event-based, local, and online computations. Yet, the state-of-the-art SNN tra…
▽ More
Spiking neural networks (SNN) are delivering energy-efficient, massively parallel, and low-latency solutions to AI problems, facilitated by the emerging neuromorphic chips. To harness these computational benefits, SNN need to be trained by learning algorithms that adhere to brain-inspired neuromorphic principles, namely event-based, local, and online computations. Yet, the state-of-the-art SNN training algorithms are based on backprop that does not follow the above principles. Due to its limited biological plausibility, the application of backprop to SNN requires non-local feedback pathways for transmitting continuous-valued errors, and relies on gradients from future timesteps. The introduction of biologically plausible modifications to backprop has helped overcome several of its limitations, but limits the degree to which backprop is approximated, which hinders its performance. We propose a biologically plausible gradient-based learning algorithm for SNN that is functionally equivalent to backprop, while adhering to all three neuromorphic principles. We introduced multi-compartment spiking neurons with local eligibility traces to compute the gradients required for learning, and a periodic "sleep" phase to further improve the approximation to backprop during which a local Hebbian rule aligns the feedback and feedforward weights. Our method achieved the same level of performance as backprop with multi-layer fully connected SNN on MNIST (98.13%) and the event-based N-MNIST (97.59%) datasets. We deployed our learning algorithm on Intel's Loihi to train a 1-hidden-layer network for MNIST, and obtained 93.32% test accuracy while consuming 400 times less energy per training sample than BioGrad on GPU. Our work shows that optimal learning is feasible in neuromorphic computing, and further pursuing its biological plausibility can better capture the benefits of this emerging computing paradigm.
△ Less
Submitted 26 October, 2021;
originally announced October 2021.
-
An Astrocyte-Modulated Neuromorphic Central Pattern Generator for Hexapod Robot Locomotion on Intel's Loihi
Authors:
Ioannis Polykretis,
Konstantinos P. Michmizos
Abstract:
Locomotion is a crucial challenge for legged robots that is addressed "effortlessly" by biological networks abundant in nature, named central pattern generators (CPG). The multitude of CPG network models that have so far become biomimetic robotic controllers is not applicable to the emerging neuromorphic hardware, depriving mobile robots of a robust walking mechanism that would result in inherentl…
▽ More
Locomotion is a crucial challenge for legged robots that is addressed "effortlessly" by biological networks abundant in nature, named central pattern generators (CPG). The multitude of CPG network models that have so far become biomimetic robotic controllers is not applicable to the emerging neuromorphic hardware, depriving mobile robots of a robust walking mechanism that would result in inherently energy-efficient systems. Here, we propose a brain-morphic CPG controler based on a comprehensive spiking neural-astrocytic network that generates two gait patterns for a hexapod robot. Building on the recently identified astrocytic mechanisms for neuromodulation, our proposed CPG architecture is seamlessly integrated into Intel's Loihi neuromorphic chip by leveraging a real-time interaction framework between the chip and the robotic operating system (ROS) environment, that we also propose. Here, we demonstrate that a Loihi-run CPG can be used to control a walking robot with robustness to sensory noise and varying speed profiles. Our results pave the way for scaling this and other approaches towards Loihi-controlled locomotion in autonomous mobile robots.
△ Less
Submitted 8 June, 2020;
originally announced June 2020.
-
Deep Learning of Movement Intent and Reaction Time for EEG-informed Adaptation of Rehabilitation Robots
Authors:
Neelesh Kumar,
Konstantinos P. Michmizos
Abstract:
Mounting evidence suggests that adaptation is a crucial mechanism for rehabilitation robots in promoting motor learning. Yet, it is commonly based on robot-derived movement kinematics, which is a rather subjective measurement of performance, especially in the presence of a sensorimotor impairment. Here, we propose a deep convolutional neural network (CNN) that uses electroencephalography (EEG) as…
▽ More
Mounting evidence suggests that adaptation is a crucial mechanism for rehabilitation robots in promoting motor learning. Yet, it is commonly based on robot-derived movement kinematics, which is a rather subjective measurement of performance, especially in the presence of a sensorimotor impairment. Here, we propose a deep convolutional neural network (CNN) that uses electroencephalography (EEG) as an objective measurement of two kinematics components that are typically used to assess motor learning and thereby adaptation: i) the intent to initiate a goal-directed movement, and ii) the reaction time (RT) of that movement. We evaluated our CNN on data acquired from an in-house experiment where 13 subjects moved a rehabilitation robotic arm in four directions on a plane, in response to visual stimuli. Our CNN achieved average test accuracies of 80.08% and 79.82% in a binary classification of the intent (intent vs. no intent) and RT (slow vs. fast), respectively. Our results demonstrate how individual movement components implicated in distinct types of motor learning can be predicted from synchronized EEG data acquired before the start of the movement. Our approach can, therefore, inform robotic adaptation in real-time and has the potential to further improve one's ability to perform the rehabilitation task.
△ Less
Submitted 18 February, 2020;
originally announced February 2020.
-
Machine Learning for Motor Learning: EEG-based Continuous Assessment of Cognitive Engagement for Adaptive Rehabilitation Robots
Authors:
Neelesh Kumar,
Konstantinos P. Michmizos
Abstract:
Although cognitive engagement (CE) is crucial for motor learning, it remains underutilized in rehabilitation robots, partly because its assessment currently relies on subjective and gross measurements taken intermittently. Here, we propose an end-to-end computational framework that assesses CE in real-time, using electroencephalography (EEG) signals as objective measurements. The framework consist…
▽ More
Although cognitive engagement (CE) is crucial for motor learning, it remains underutilized in rehabilitation robots, partly because its assessment currently relies on subjective and gross measurements taken intermittently. Here, we propose an end-to-end computational framework that assesses CE in real-time, using electroencephalography (EEG) signals as objective measurements. The framework consists of i) a deep convolutional neural network (CNN) that extracts task-discriminative spatiotemporal EEG to predict the level of CE for two classes -- cognitively engaged vs. disengaged; and ii) a novel sliding window method that predicts continuous levels of CE in real-time. We evaluated our framework on 8 subjects using an in-house Go/No-Go experiment that adapted its gameplay parameters to induce cognitive fatigue. The proposed CNN had an average leave-one-out accuracy of 88.13\%. The CE prediction correlated well with a commonly used behavioral metric based on self-reports taken every 5 minutes ($ρ$=0.93). Our results objectify CE in real-time and pave the way for using CE as a rehabilitation parameter for tailoring robotic therapy to each patient's needs and skills.
△ Less
Submitted 19 February, 2020; v1 submitted 18 February, 2020;
originally announced February 2020.
-
A Spiking Neural Network Emulating the Structure of the Oculomotor System Requires No Learning to Control a Biomimetic Robotic Head
Authors:
Praveenram Balachandar,
Konstantinos P. Michmizos
Abstract:
Robotic vision introduces requirements for real-time processing of fast-varying, noisy information in a continuously changing environment. In a real-world environment, convenient assumptions, such as static camera systems and deep learning algorithms devouring high volumes of ideally slightly-varying data are hard to survive. Leveraging on recent studies on the neural connectome associated with ey…
▽ More
Robotic vision introduces requirements for real-time processing of fast-varying, noisy information in a continuously changing environment. In a real-world environment, convenient assumptions, such as static camera systems and deep learning algorithms devouring high volumes of ideally slightly-varying data are hard to survive. Leveraging on recent studies on the neural connectome associated with eye movements, we designed a neuromorphic oculomotor controller and placed it at the heart of our in-house biomimetic robotic head prototype. The controller is unique in the sense that (1) all data are encoded and processed by a spiking neural network (SNN), and (2) by mimicking the associated brain areas' topology, the SNN is biologically interpretable and requires no training to operate. Here, we report the robot's target tracking ability, demonstrate that its eye kinematics are similar to those reported in human eye studies and show that a biologically-constrained learning, although not required for the SNN's function, can be used to further refine its performance. This work aligns with our ongoing effort to develop energy-efficient neuromorphic SNNs and harness their emerging intelligence to control biomimetic robots with versatility and robustness.
△ Less
Submitted 8 June, 2020; v1 submitted 18 February, 2020;
originally announced February 2020.
-
Introducing Astrocytes on a Neuromorphic Processor: Synchronization, Local Plasticity and Edge of Chaos
Authors:
Guangzhi Tang,
Ioannis E. Polykretis,
Vladimir A. Ivanov,
Arpit Shah,
Konstantinos P. Michmizos
Abstract:
While there is still a lot to learn about astrocytes and their neuromodulatory role in the spatial and temporal integration of neuronal activity, their introduction to neuromorphic hardware is timely, facilitating their computational exploration in basic science questions as well as their exploitation in real-world applications. Here, we present an astrocytic module that enables the development of…
▽ More
While there is still a lot to learn about astrocytes and their neuromodulatory role in the spatial and temporal integration of neuronal activity, their introduction to neuromorphic hardware is timely, facilitating their computational exploration in basic science questions as well as their exploitation in real-world applications. Here, we present an astrocytic module that enables the development of a spiking Neuronal-Astrocytic Network (SNAN) into Intel's Loihi neuromorphic chip. The basis of the Loihi module is an end-to-end biophysically plausible compartmental model of an astrocyte that simulates the intracellular activity in response to the synaptic activity in space and time. To demonstrate the functional role of astrocytes in SNAN, we describe how an astrocyte may sense and induce activity-dependent neuronal synchronization, switch on and off spike-time-dependent plasticity (STDP) to introduce single-shot learning, and monitor the transition between ordered and chaotic activity at the synaptic space. Our module may serve as an extension for neuromorphic hardware, by either replicating or exploring the distinct computational roles that astrocytes have in forming biological intelligence.
△ Less
Submitted 19 September, 2019; v1 submitted 2 July, 2019;
originally announced July 2019.
-
Axonal Conduction Velocity Impacts Neuronal Network Oscillations
Authors:
Vladimir A. Ivanov,
Ioannis E. Polykretis,
Konstantinos P. Michmizos
Abstract:
Increasing experimental evidence suggests that axonal action potential conduction velocity is a highly adaptive parameter in the adult central nervous system. Yet, the effects of this newfound plasticity on global brain dynamics is poorly understood. In this work, we analyzed oscillations in biologically plausible neuronal networks with different conduction velocity distributions. Changes of 1-2 (…
▽ More
Increasing experimental evidence suggests that axonal action potential conduction velocity is a highly adaptive parameter in the adult central nervous system. Yet, the effects of this newfound plasticity on global brain dynamics is poorly understood. In this work, we analyzed oscillations in biologically plausible neuronal networks with different conduction velocity distributions. Changes of 1-2 (ms) in network mean signal transmission time resulted in substantial network oscillation frequency changes ranging in 0-120 (Hz). Our results suggest that changes in axonal conduction velocity may significantly affect both the frequency and synchrony of brain rhythms, which have well established connections to learning, memory, and other cognitive processes.
△ Less
Submitted 22 March, 2019;
originally announced March 2019.
-
Computational Astrocyence: Astrocytes encode inhibitory activity into the frequency and spatial extent of their calcium elevations
Authors:
Ioannis E. Polykretis,
Vladimir A. Ivanov,
Konstantinos P. Michmizos
Abstract:
Deciphering the complex interactions between neurotransmission and astrocytic $Ca^{2+}$ elevations is a target promising a comprehensive understanding of brain function. While the astrocytic response to excitatory synaptic activity has been extensively studied, how inhibitory activity results to intracellular $Ca^{2+}$ waves remains elusive. In this study, we developed a compartmental astrocytic m…
▽ More
Deciphering the complex interactions between neurotransmission and astrocytic $Ca^{2+}$ elevations is a target promising a comprehensive understanding of brain function. While the astrocytic response to excitatory synaptic activity has been extensively studied, how inhibitory activity results to intracellular $Ca^{2+}$ waves remains elusive. In this study, we developed a compartmental astrocytic model that exhibits distinct levels of responsiveness to inhibitory activity. Our model suggested that the astrocytic coverage of inhibitory terminals defines the spatial and temporal scale of their $Ca^{2+}$ elevations. Understanding the interplay between the synaptic pathways and the astrocytic responses will help us identify how astrocytes work independently and cooperatively with neurons, in health and disease.
△ Less
Submitted 18 March, 2019;
originally announced March 2019.
-
A Neural-Astrocytic Network Architecture: Astrocytic calcium waves modulate synchronous neuronal activity
Authors:
Ioannis Polykretis,
Vladimir Ivanov,
Konstantinos P. Michmizos
Abstract:
Understanding the role of astrocytes in brain computation is a nascent challenge, promising immense rewards, in terms of new neurobiological knowledge that can be translated into artificial intelligence. In our ongoing effort to identify principles endow-ing the astrocyte with unique functions in brain computation, and translate them into neural-astrocytic networks (NANs), we propose a biophysical…
▽ More
Understanding the role of astrocytes in brain computation is a nascent challenge, promising immense rewards, in terms of new neurobiological knowledge that can be translated into artificial intelligence. In our ongoing effort to identify principles endow-ing the astrocyte with unique functions in brain computation, and translate them into neural-astrocytic networks (NANs), we propose a biophysically realistic model of an astrocyte that preserves the experimentally observed spatial allocation of its distinct subcellular compartments. We show how our model may encode, and modu-late, the extent of synchronous neural activity via calcium waves that propagate intracellularly across the astrocytic compartments. This relationship between neural activity and astrocytic calcium waves has long been speculated but it is still lacking a mechanistic explanation. Our model suggests an astrocytic "calcium cascade" mechanism for neuronal synchronization, which may empower NANs by imposing periodic neural modulation known to reduce coding errors. By expanding our notions of information processing in astrocytes, our work aims to solidify a computational role for non-neuronal cells and incorporate them into artificial networks.
△ Less
Submitted 5 July, 2018;
originally announced July 2018.
-
Gridbot: An autonomous robot controlled by a Spiking Neural Network mimicking the brain's navigational system
Authors:
Guangzhi Tang,
Konstantinos P. Michmizos
Abstract:
It is true that the "best" neural network is not necessarily the one with the most "brain-like" behavior. Understanding biological intelligence, however, is a fundamental goal for several distinct disciplines. Translating our understanding of intelligence to machines is a fundamental problem in robotics. Propelled by new advancements in Neuroscience, we developed a spiking neural network (SNN) tha…
▽ More
It is true that the "best" neural network is not necessarily the one with the most "brain-like" behavior. Understanding biological intelligence, however, is a fundamental goal for several distinct disciplines. Translating our understanding of intelligence to machines is a fundamental problem in robotics. Propelled by new advancements in Neuroscience, we developed a spiking neural network (SNN) that draws from mounting experimental evidence that a number of individual neurons is associated with spatial navigation. By following the brain's structure, our model assumes no initial all-to-all connectivity, which could inhibit its translation to a neuromorphic hardware, and learns an uncharted territory by mapping its identified components into a limited number of neural representations, through spike-timing dependent plasticity (STDP). In our ongoing effort to employ a bioinspired SNN-controlled robot to real-world spatial mapping applications, we demonstrate here how an SNN may robustly control an autonomous robot in mapping and exploring an unknown environment, while compensating for its own intrinsic hardware imperfections, such as partial or total loss of visual input.
△ Less
Submitted 5 July, 2018;
originally announced July 2018.
-
Maturation Trajectories of Cortical Resting-State Networks Depend on the Mediating Frequency Band
Authors:
Sheraz Khan,
Javeria Hashmi,
Fahimeh Mamashli,
Konstantinos Michmizos,
Manfred Kitzbichler,
Hari Bharadwaj,
Yousra Bekhti,
Santosh Ganesan,
Keri A Garel,
Susan Whitfield-Gabrieli,
Randy Gollub,
Jian Kong,
Lucia M Vaina,
Kunjan Rana,
Steven Stufflebeam,
Matti Hamalainen,
Tal Kenet
Abstract:
The functional significance of resting state networks and their abnormal manifestations in psychiatric disorders are firmly established, as is the importance of the cortical rhythms in mediating these networks. Resting state networks are known to undergo substantial reorganization from childhood to adulthood, but whether distinct cortical rhythms, which are generated by separable neural mechanisms…
▽ More
The functional significance of resting state networks and their abnormal manifestations in psychiatric disorders are firmly established, as is the importance of the cortical rhythms in mediating these networks. Resting state networks are known to undergo substantial reorganization from childhood to adulthood, but whether distinct cortical rhythms, which are generated by separable neural mechanisms and are often manifested abnormally in psychiatric conditions, mediate maturation differentially, remains unknown. Using magnetoencephalography (MEG) to map frequency band specific maturation of resting state networks from age 7 to 29 in 162 participants (31 independent), we found significant changes with age in networks mediated by the beta (13-30Hz) and gamma (31-80Hz) bands. More specifically, gamma band mediated networks followed an expected asymptotic trajectory, but beta band mediated networks followed a linear trajectory. Network integration increased with age in gamma band mediated networks, while local segregation increased with age in beta band mediated networks. Spatially, the hubs that changed in importance with age in the beta band mediated networks had relatively little overlap with those that showed the greatest changes in the gamma band mediated networks. These findings are relevant for our understanding of the neural mechanisms of cortical maturation, in both typical and atypical development.
△ Less
Submitted 12 February, 2018;
originally announced March 2018.
-
Sequence learning in Associative Neuronal-Astrocytic Network
Authors:
Leo Kozachkov,
Konstantinos P. Michmizos
Abstract:
The neuronal paradigm of studying the brain has left us with limitations in both our understanding of how neurons process information to achieve biological intelligence and how such knowledge may be translated into artificial intelligence and its most brain-derived branch, neuromorphic computing. Overturning our fundamental assumptions of how the brain works, the recent exploration of astrocytes i…
▽ More
The neuronal paradigm of studying the brain has left us with limitations in both our understanding of how neurons process information to achieve biological intelligence and how such knowledge may be translated into artificial intelligence and its most brain-derived branch, neuromorphic computing. Overturning our fundamental assumptions of how the brain works, the recent exploration of astrocytes is revealing that these long-neglected brain cells dynamically regulate learning by interacting with neuronal activity at the synaptic level. Following recent experimental evidence, we designed an associative, Hopfield-type, neuronal-astrocytic network and analyzed the dynamics of the interaction between neurons and astrocytes. We show that astrocytes were sufficient to trigger transitions between learned memories in the neuronal component of the network. Further, we mathematically derived the timing of the transitions that was governed by the dynamics of the calcium-dependent slow-currents in the astrocytic processes. Overall, we provide a brain-morphic mechanism for sequence learning that is inspired by, and aligns with, recent experimental findings. To evaluate our model, we emulated astrocytic atrophy and showed that memory recall becomes significantly impaired after a critical point of affected astrocytes was reached. This brain-inspired and brain-validated approach supports our ongoing efforts to incorporate non-neuronal computing elements in neuromorphic information processing.
△ Less
Submitted 10 May, 2020; v1 submitted 16 July, 2017;
originally announced July 2017.
-
The Causal Role of Astrocytes in Slow-Wave Rhythmogenesis: A Computational Modelling Study
Authors:
Leo Kozachkov,
Konstantinos P. Michmizos
Abstract:
Finding the origin of slow and infra-slow oscillations could reveal or explain brain mechanisms in health and disease. Here, we present a biophysically constrained computational model of a neural network where the inclusion of astrocytes introduced slow and infra-slow-oscillations, through two distinct mechanisms. Specifically, we show how astrocytes can modulate the fast network activity through…
▽ More
Finding the origin of slow and infra-slow oscillations could reveal or explain brain mechanisms in health and disease. Here, we present a biophysically constrained computational model of a neural network where the inclusion of astrocytes introduced slow and infra-slow-oscillations, through two distinct mechanisms. Specifically, we show how astrocytes can modulate the fast network activity through their slow inter-cellular calcium wave speed and amplitude and possibly cause the oscillatory imbalances observed in diseases commonly known for such abnormalities, namely Alzheimer's disease, Parkinson's disease, epilepsy, depression and ischemic stroke. This work aims to increase our knowledge on how astrocytes and neurons synergize to affect brain function and dysfunction.
△ Less
Submitted 13 February, 2017;
originally announced February 2017.
