Introducing the Dendrify framework for incorporating dendrites to spiking neural networks

The wave nature of the action potential

An alternative to the standard Hodgkin-Huxley model for the action potential in axons is presented. It is based on our recently developed theory of electric field wave propagation in anisotropic and inhomogeneous brain tissues, which has been shown to explain a broad range of observed coherent synchronous brain electrical processes. We demonstrate that this theory also explains the spiking behavior of single neurons, thereby bridging the gap between the fundamental element of brain electrical activity-the neuron-and large-scale coherent synchronous electrical activity. We demonstrate that our recently developed theory of electric field wave propagation in anisotropic and inhomogeneous brain tissues, which has been shown to explain a broad range of observed coherent synchronous brain electrical processes, also applies to the spiking behavior of single neurons, thus bridging the gap between the fundamental element of brain electrical activity (the neuron) and large-scale coherent synchronous electrical activity. Our analysis indicates that a non-linear system with several small parameters can mathematically describe the membrane interface of the axonal cellular system. This enables the rigorous derivation of an accurate yet simpler non-linear model through the formal small-parameter expansion. The resulting action potential model exhibits a smooth, continuous transition from the linear wave oscillatory regime to the non-linear spiking regime, as well as a critical transition to a non-oscillatory regime. These transitions occur with changes in the criticality parameter and include several different bifurcation types, representative of the various experimentally detected neuron types. This new theory addresses the limitations of the Hodgkin-Huxley model, including its inability to explain extracellular spiking, efficient brain synchronization, saltatory conduction along myelinated axons, and various other observed coherent macroscopic brain electrical phenomena. We also demonstrate that our approach recovers the standard cable axon theory, utilizing the relatively simple assumptions of piece-wise homogeneity and isotropy. However, the diffusion process described by the cable equation is not capable of supporting action potential propagation across a wide range of experimentally reported axon parameters.

Cardiac Heterogeneity Prediction by Cardio-Neural Network Simulation

The bidirectional interactions between brain and heart through autonomic nervous system is the prime focus of neuro-cardiology community. The computer models designed to analyze brain and heart signals are either complex in terms of molecular and cellular interactions or not capable of representing the complex ion channel dynamics. Therefore, scientists are unable to extract the overall behavior of organs by electrical response of heterogeneous cells of brain and heart. In this study, a unified model of excitable cells is proposed that can be modulated by adrenergic features. By implementing the proposed model, a network of one thousand sparsely coupled cardio-neural network is simulated. The major findings of study include i. cardiac heterogeneity in electrical behavior of cardiac myocytes is the prime factor of heart rate variability ii. Brain-heart interplay through electrical pulses holds the necessary information of brain and heart signals that can be analyzed through spiking neural networks iii. Heart rate variability can be predicted and monitored by spiking neural networks from electrophysiological recordings of brain and heart iv. Heart rate variability related to tachycardia and bradycardia depends upon the polarization protocols of cardiac myocytes during plateau phase of action potential. This study provides the modeling and simulation phase of brain-heart interface to predict the morbidity at early stages. The recent advancements in nano-electronics will make is possible to develop brain-heart interface as nano-chip to deploy in subject to stimulate the brain-heart interplay through electrophysiological signals.

Highly Reliable Bi2O2Se Dendritic Neuron Enabling Spatial-Temporal Signal Processing for Real-World Image Classification

Artificial intelligence (AI) has made significant strides by imitating biological neurons and synapses through simplified models, yet incomplete neuron functionalities can limit performance and energy efficiency in handling complex tasks. Biological neurons process input signals nonlinearly, utilizing dendrites to process spatial-temporal information. This study demonstrates the compact artificial dendrite device employing memristors based on bismuth oxyselenide (Bi2O2Se). Transfer-free Bi2O2Se switching medium is directly grown on the metal-patterned substrates via 350 °C selenization process. The layered Bi2O2Se structure, limiting metal injection, results in reliable dynamic resistive switching with excellent cycle uniformity and exceptional endurance over 2 million cycles. The highly reliable current response of dynamic resistive switching is modeled with respect to the spatial-temporal voltage input. With the Bi2O2Se dendrite device, dendritic neuron model is implemented, and the proposed neural network achieved high recognition rates of 78.3% with the street view house numbers (SVHN) data set.

The calcitron: A simple neuron model that implements many learning rules via the calcium control hypothesis

Theoretical neuroscientists and machine learning researchers have proposed a variety of learning rules to enable artificial neural networks to effectively perform both supervised and unsupervised learning tasks. It is not always clear, however, how these theoretically-derived rules relate to biological mechanisms of plasticity in the brain, or how these different rules might be mechanistically implemented in different contexts and brain regions. This study shows that the calcium control hypothesis, which relates synaptic plasticity in the brain to the calcium concentration ([Ca2+]) in dendritic spines, can produce a diverse array of learning rules. We propose a simple, perceptron-like neuron model, the calcitron, that has four sources of [Ca2+]: local (following the activation of an excitatory synapse and confined to that synapse), heterosynaptic (resulting from the activity of other synapses), postsynaptic spike-dependent, and supervisor-dependent. We demonstrate that by modulating the plasticity thresholds and calcium influx from each calcium source, we can reproduce a wide range of learning and plasticity protocols, such as Hebbian and anti-Hebbian learning, frequency-dependent plasticity, and unsupervised recognition of frequently repeating input patterns. Moreover, by devising simple neural circuits to provide supervisory signals, we show how the calcitron can implement homeostatic plasticity, perceptron learning, and BTSP-inspired one-shot learning. Our study bridges the gap between theoretical learning algorithms and their biological counterparts, not only replicating established learning paradigms but also introducing novel rules.

Of mice and men: Dendritic architecture differentiates human from mice neuronal networks

The organizational principles that distinguish the human brain from other species have been a long-standing enigma in neuroscience. Focusing on the uniquely evolved human cortical layers 2 and 3, we computationally reconstruct the cortical architecture for mice and humans. We show that human pyramidal cells form highly complex networks, demonstrated by the increased number and simplex dimension compared to mice. This is surprising because human pyramidal cells are much sparser in the cortex. We show that the number and size of neurons fail to account for this increased network complexity, suggesting that another morphological property is a key determinant of network connectivity. Topological comparison of dendritic structure reveals much higher perisomatic (basal and oblique) branching density in human pyramidal cells. Using topological tools we quantitatively show that this neuronal structural property directly impacts network complexity, including the formation of a rich subnetwork structure. We conclude that greater dendritic complexity, a defining attribute of human L2 and 3 neurons, may provide the human cortex with enhanced computational capacity and cognitive flexibility.

Multi-scale modeling to investigate the effects of transcranial magnetic stimulation on morphologically-realistic neuron with depression

Transcranial magnetic stimulation (TMS) is a non-invasive neuromodulation technique to activate or inhibit the activity of neurons, and thereby regulate their excitability. This technique has demonstrated potential in the treatment of neuropsychiatric disorders, such as depression. However, the effect of TMS on neurons with different severity of depression is still unclear, limiting the development of efficient and personalized clinical application parameters. In this study, a multi-scale computational model was developed to investigate and quantify the differences in neuronal responses to TMS with different degrees of depression. The microscale neuronal models we constructed represent the hippocampal CA1 region in rats under normal conditions and with varying severities of depression (mild, moderate, and major depressive disorder). These models were then coupled to a macroscopic TMS-induced E-Fields model of a rat head comprising multiple types of tissue. Our results demonstrate alterations in neuronal membrane potential and calcium concentration across varying levels of depression severity. As depression severity increases, the peak membrane potential and polarization degree of neuronal soma and dendrites gradually decline, while the peak calcium concentration decreases and the peak arrival time prolongs. Concurrently, the electric fields thresholds and amplification coefficient gradually rise, indicating an increasing difficulty in activating neurons with depression. This study offers novel insights into the mechanisms of magnetic stimulation in depression treatment using multi-scale computational models. It underscores the importance of considering depression severity in treatment strategies, promising to optimize TMS therapeutic approaches.

DendroTweaks: An interactive approach for unraveling dendritic dynamics

Neurons rely on the interplay between dendritic morphology and ion channels to transform synaptic inputs into a sequence of somatic spikes. Detailed biophysical models with active dendrites have been instrumental in exploring this interaction. However, such models can be challenging to understand and validate due to the large number of parameters involved. In this work, we introduce DendroTweaks - a toolbox designed to illuminate how morpho-electric properties map to dendritic events and how these dendritic events shape neuronal output. DendroTweaks features a web-based graphical interface, where users can explore single-cell neuronal models and adjust their morphological and biophysical parameters with real-time visual feedback. In particular, DendroTweaks is tailored to interactive fine-tuning of subcellular properties, such as kinetics and distributions of ion channels, as well as the dynamics and allocation of synaptic inputs. It offers an automated approach for standardization and refinement of voltage-gated ion channel models to make them more comprehensible and reusable. The toolbox allows users to run various experimental protocols and record data from multiple dendritic and somatic locations, thereby enhancing model validation. Finally, it aims to deepen our understanding of which dendritic properties are essential for neuronal input-output transformation. Using this knowledge, one can simplify models through a built-in morphology reduction algorithm and export them for further use in faster, more interpretable networks. With DendroTweaks, users can gain better control and understanding of their models, advancing research on dendritic input-output transformations and their role in network computations.

Memory-efficient neurons and synapses for spike-timing-dependent-plasticity in large-scale spiking networks

This paper addresses the challenges posed by frequent memory access during simulations of large-scale spiking neural networks involving synaptic plasticity. We focus on the memory accesses performed during a common synaptic plasticity rule since this can be a significant factor limiting the efficiency of the simulations. We propose neuron models that are represented by only three state variables, which are engineered to enforce the appropriate neuronal dynamics. Additionally, memory retrieval is executed solely by fetching postsynaptic variables, promoting a contiguous memory storage and leveraging the capabilities of burst mode operations to reduce the overhead associated with each access. Different plasticity rules could be implemented despite the adopted simplifications, each leading to a distinct synaptic weight distribution (i.e., unimodal and bimodal). Moreover, our method requires fewer average memory accesses compared to a naive approach. We argue that the strategy described can speed up memory transactions and reduce latencies while maintaining a small memory footprint.

Bridging model and experiment in systems neuroscience with Cleo: the Closed-Loop, Electrophysiology, and Optophysiology simulation testbed

Systems neuroscience has experienced an explosion of new tools for reading and writing neural activity, enabling exciting new experiments such as all-optical or closed-loop control that effect powerful causal interventions. At the same time, improved computational models are capable of reproducing behavior and neural activity with increasing fidelity. Unfortunately, these advances have drastically increased the complexity of integrating different lines of research, resulting in the missed opportunities and untapped potential of suboptimal experiments. Experiment simulation can help bridge this gap, allowing model and experiment to better inform each other by providing a low-cost testbed for experiment design, model validation, and methods engineering. Specifically, this can be achieved by incorporating the simulation of the experimental interface into our models, but no existing tool integrates optogenetics, two-photon calcium imaging, electrode recording, and flexible closed-loop processing with neural population simulations. To address this need, we have developed Cleo: the Closed-Loop, Electrophysiology, and Optophysiology experiment simulation testbed. Cleo is a Python package enabling injection of recording and stimulation devices as well as closed-loop control with realistic latency into a Brian spiking neural network model. It is the only publicly available tool currently supporting two-photon and multi-opsin/wavelength optogenetics. To facilitate adoption and extension by the community, Cleo is open-source, modular, tested, and documented, and can export results to various data formats. Here we describe the design and features of Cleo, validate output of individual components and integrated experiments, and demonstrate its utility for advancing optogenetic techniques in prospective experiments using previously published systems neuroscience models.

Network motifs in cellular neurophysiology

Concepts from network science and graph theory, including the framework of network motifs, have been frequently applied in studying neuronal networks and other biological complex systems. Network-based approaches can also be used to study the functions of individual neurons, where cellular elements such as ion channels and membrane voltage are conceptualized as nodes within a network, and their interactions are denoted by edges. Network motifs in this context provide functional building blocks that help to illuminate the principles of cellular neurophysiology. In this review we build a case that network motifs operating within neurons provide tools for defining the functional architecture of single-neuron physiology and neuronal adaptations. We highlight the presence of such computational motifs in the cellular mechanisms underlying action potential generation, neuronal oscillations, dendritic integration, and neuronal plasticity. Future work applying the network motifs perspective may help to decipher the functional complexities of neurons and their adaptation during health and disease.

The switching and learning behavior of an octopus cell implemented on FPGA

A dendrocentric backpropagation spike timing-dependent plasticity learning rule has been derived based on temporal logic for a single octopus neuron. It receives parallel spike trains and collectively adjusts its synaptic weights in the range [0, 1] during training. After the training phase, it spikes in reaction to event signaling input patterns in sensory streams. The learning and switching behavior of the octopus cell has been implemented in field-programmable gate array (FPGA) hardware. The application in an FPGA is described and the proof of concept for its application in hardware that was obtained by feeding it with spike cochleagrams is given; also, it is verified by performing a comparison with the pre-computed standard software simulation results.

Leveraging dendritic properties to advance machine learning and neuro-inspired computing

The brain is a remarkably capable and efficient system. It can process and store huge amounts of noisy and unstructured information, using minimal energy. In contrast, current artificial intelligence (AI) systems require vast resources for training while still struggling to compete in tasks that are trivial for biological agents. Thus, brain-inspired engineering has emerged as a promising new avenue for designing sustainable, next-generation AI systems. Here, we describe how dendritic mechanisms of biological neurons have inspired innovative solutions for significant AI problems, including credit assignment in multi-layer networks, catastrophic forgetting, and high-power consumption. These findings provide exciting alternatives to existing architectures, showing how dendritic research can pave the way for building more powerful and energy efficient artificial learning systems.

Network-neuron interactions underlying sensory responses of layer 5 pyramidal tract neurons in barrel cortex

Neurons in the cerebral cortex receive thousands of synaptic inputs per second from thousands of presynaptic neurons. How the dendritic location of inputs, their timing, strength, and presynaptic origin, in conjunction with complex dendritic physiology, impact the transformation of synaptic input into action potential (AP) output remains generally unknown for in vivo conditions. Here, we introduce a computational approach to reveal which properties of the input causally underlie AP output, and how this neuronal input-output computation is influenced by the morphology and biophysical properties of the dendrites. We demonstrate that this approach allows dissecting of how different input populations drive in vivo observed APs. For this purpose, we focus on fast and broadly tuned responses that pyramidal tract neurons in layer 5 (L5PTs) of the rat barrel cortex elicit upon passive single whisker deflections. By reducing a multi-scale model that we reported previously, we show that three features are sufficient to predict with high accuracy the sensory responses and receptive fields of L5PTs under these specific in vivo conditions: the count of active excitatory versus inhibitory synapses preceding the response, their spatial distribution on the dendrites, and the AP history. Based on these three features, we derive an analytically tractable description of the input-output computation of L5PTs, which enabled us to dissect how synaptic input from thalamus and different cell types in barrel cortex contribute to these responses. We show that the input-output computation is preserved across L5PTs despite morphological and biophysical diversity of their dendrites. We found that trial-to-trial variability in L5PT responses, and cell-to-cell variability in their receptive fields, are sufficiently explained by variability in synaptic input from the network, whereas variability in biophysical and morphological properties have minor contributions. Our approach to derive analytically tractable models of input-output computations in L5PTs provides a roadmap to dissect network-neuron interactions underlying L5PT responses across different in vivo conditions and for other cell types.

Temporal dendritic heterogeneity incorporated with spiking neural networks for learning multi-timescale dynamics

It is widely believed the brain-inspired spiking neural networks have the capability of processing temporal information owing to their dynamic attributes. However, how to understand what kind of mechanisms contributing to the learning ability and exploit the rich dynamic properties of spiking neural networks to satisfactorily solve complex temporal computing tasks in practice still remains to be explored. In this article, we identify the importance of capturing the multi-timescale components, based on which a multi-compartment spiking neural model with temporal dendritic heterogeneity, is proposed. The model enables multi-timescale dynamics by automatically learning heterogeneous timing factors on different dendritic branches. Two breakthroughs are made through extensive experiments: the working mechanism of the proposed model is revealed via an elaborated temporal spiking XOR problem to analyze the temporal feature integration at different levels; comprehensive performance benefits of the model over ordinary spiking neural networks are achieved on several temporal computing benchmarks for speech recognition, visual recognition, electroencephalogram signal recognition, and robot place recognition, which shows the best-reported accuracy and model compactness, promising robustness and generalization, and high execution efficiency on neuromorphic hardware. This work moves neuromorphic computing a significant step toward real-world applications by appropriately exploiting biological observations.

Action at a Distance: Theoretical Mechanisms of Cross-Dendritic Heterosynaptic Modification

Highlighted Research Paper: T. Moldwin, M. Kalmenson, and I. Segev, "Asymmetric voltage attenuation in dendrites can enable hierarchical heterosynaptic plasticity." eNeuro (2023).

Distinctive properties of biological neural networks and recent advances in bottom-up approaches toward a better biologically plausible neural network

Although it may appear infeasible and impractical, building artificial intelligence (AI) using a bottom-up approach based on the understanding of neuroscience is straightforward. The lack of a generalized governing principle for biological neural networks (BNNs) forces us to address this problem by converting piecemeal information on the diverse features of neurons, synapses, and neural circuits into AI. In this review, we described recent attempts to build a biologically plausible neural network by following neuroscientifically similar strategies of neural network optimization or by implanting the outcome of the optimization, such as the properties of single computational units and the characteristics of the network architecture. In addition, we proposed a formalism of the relationship between the set of objectives that neural networks attempt to achieve, and neural network classes categorized by how closely their architectural features resemble those of BNN. This formalism is expected to define the potential roles of top-down and bottom-up approaches for building a biologically plausible neural network and offer a map helping the navigation of the gap between neuroscience and AI engineering.

Leveraging dendritic properties to advance machine learning and neuro-inspired computing

The brain is a remarkably capable and efficient system. It can process and store huge amounts of noisy and unstructured information using minimal energy. In contrast, current artificial intelligence (AI) systems require vast resources for training while still struggling to compete in tasks that are trivial for biological agents. Thus, brain-inspired engineering has emerged as a promising new avenue for designing sustainable, next-generation AI systems. Here, we describe how dendritic mechanisms of biological neurons have inspired innovative solutions for significant AI problems, including credit assignment in multilayer networks, catastrophic forgetting, and high energy consumption. These findings provide exciting alternatives to existing architectures, showing how dendritic research can pave the way for building more powerful and energy-efficient artificial learning systems.