Advances in memristor based artificial neuron fabrication-materials,models,and applications

Spiking neural network(SNN),widely known as the third-generation neural network,has been frequently investigated due to its excellent spatiotemporal information processing capability,high biological plausibility,and low energy consumption characteristics.Analogous to the working mechanism of human brain,the SNN system transmits information through the spiking action of neurons.Therefore,artificial neurons are critical building blocks for constructing SNN in hardware.Memristors are drawing growing attention due to low consumption,high speed,and nonlinearity characteristics,which are recently introduced to mimic the functions of biological neurons.Researchers have proposed multifarious memristive materials including organic materials,inorganic materials,or even two-dimensional materials.Taking advantage of the unique electrical behavior of these materials,several neuron models are successfully implemented,such as Hodgkin–Huxley model,leaky integrate-and-fire model and integrate-and-fire model.In this review,the recent reports of artificial neurons based on memristive devices are discussed.In addition,we highlight the models and applications through combining artificial neuronal devices with sensors or other electronic devices.Finally,the future challenges and outlooks of memristor-based artificial neurons are discussed,and the development of hardware implementation of brain-like intelligence system based on SNN is also prospected.

Dynamics and synchronization of neural models with memristive membranes under energy coupling

Dynamical modeling of neural systems plays an important role in explaining and predicting some features of biophysical mechanisms.The electrophysiological environment inside and outside of the nerve cell is different.Due to the continuous and periodical properties of electromagnetic fields in the cell during its operation,electronic components involving two capacitors and a memristor are effective in mimicking these physical features.In this paper,a neural circuit is reconstructed by two capacitors connected by a memristor with periodical mem-conductance.It is found that the memristive neural circuit can present abundant firing patterns without stimulus.The Hamilton energy function is deduced using the Helmholtz theorem.Further,a neuronal network consisting of memristive neurons is proposed by introducing energy coupling.The controllability and flexibility of parameters give the model the ability to describe the dynamics and synchronization behavior of the system.

Exploiting fly models to investigate rare human neurological disorders

Rare neurological diseases,while individually are rare,collectively impact millions globally,leading to diverse and often severe neurological symptoms.Often attributed to genetic mutations that disrupt protein function or structure,understanding their genetic basis is crucial for accurate diagnosis and targeted therapies.To investigate the underlying pathogenesis of these conditions,researchers often use non-mammalian model organisms,such as Drosophila(fruit flies),which is valued for their genetic manipulability,cost-efficiency,and preservation of genes and biological functions across evolutionary time.Genetic tools available in Drosophila,including CRISPR-Cas9,offer a means to manipulate gene expression,allowing for a deep exploration of the genetic underpinnings of rare neurological diseases.Drosophila boasts a versatile genetic toolkit,rapid generation turnover,and ease of large-scale experimentation,making it an invaluable resource for identifying potential drug candidates.Researchers can expose flies carrying disease-associated mutations to various compounds,rapidly pinpointing promising therapeutic agents for further investigation in mammalian models and,ultimately,clinical trials.In this comprehensive review,we explore rare neurological diseases where fly research has significantly contributed to our understanding of their genetic basis,pathophysiology,and potential therapeutic implications.We discuss rare diseases associated with both neuron-expressed and glial-expressed genes.Specific cases include mutations in CDK19 resulting in epilepsy and developmental delay,mutations in TIAM1 leading to a neurodevelopmental disorder with seizures and language delay,and mutations in IRF2BPL causing seizures,a neurodevelopmental disorder with regression,loss of speech,and abnormal movements.And we explore mutations in EMC1 related to cerebellar atrophy,visual impairment,psychomotor retardation,and gain-of-function mutations in ACOX1 causing Mitchell syndrome.Loss-of-function mutations in ACOX1 result in ACOX1 deficiency,characterized by very-long-chain fatty acid accumulation and glial degeneration.Notably,this review highlights how modeling these diseases in Drosophila has provided valuable insights into their pathophysiology,offering a platform for the rapid identification of potential therapeutic interventions.Rare neurological diseases involve a wide range of expression systems,and sometimes common phenotypes can be found among different genes that cause abnormalities in neurons or glia.Furthermore,mutations within the same gene may result in varying functional outcomes,such as complete loss of function,partial loss of function,or gain-of-function mutations.The phenotypes observed in patients can differ significantly,underscoring the complexity of these conditions.In conclusion,Drosophila represents an indispensable and cost-effective tool for investigating rare neurological diseases.By facilitating the modeling of these conditions,Drosophila contributes to a deeper understanding of their genetic basis,pathophysiology,and potential therapies.This approach accelerates the discovery of promising drug candidates,ultimately benefiting patients affected by these complex and understudied diseases.

The burden of upper motor neuron involvement is correlated with the bilateral limb involvement interval in patients with amyotrophic lateral sclerosis:a retrospective observational study

Amyotrophic lateral sclerosis is a rare neurodegenerative disease characterized by the involvement of both upper and lower motor neurons.Early bilateral limb involvement significantly affects patients'daily lives and may lead them to be confined to bed.However,the effect of upper and lower motor neuron impairment and other risk factors on bilateral limb involvement is unclear.To address this issue,we retrospectively collected data from 586 amyotrophic lateral sclerosis patients with limb onset diagnosed at Peking University Third Hospital between January 2020 and May 2022.A univariate analysis revealed no significant differences in the time intervals of spread in different directions between individuals with upper motor neuron-dominant amyotrophic lateral sclerosis and those with classic amyotrophic lateral sclerosis.We used causal directed acyclic graphs for risk factor determination and Cox proportional hazards models to investigate the association between the duration of bilateral limb involvement and clinical baseline characteristics in amyotrophic lateral sclerosis patients.Multiple factor analyses revealed that higher upper motor neuron scores(hazard ratio[HR]=1.05,95%confidence interval[CI]=1.01–1.09,P=0.018),onset in the left limb(HR=0.72,95%CI=0.58–0.89,P=0.002),and a horizontal pattern of progression(HR=0.46,95%CI=0.37–0.58,P<0.001)were risk factors for a shorter interval until bilateral limb involvement.The results demonstrated that a greater degree of upper motor neuron involvement might cause contralateral limb involvement to progress more quickly in limb-onset amyotrophic lateral sclerosis patients.These findings may improve the management of amyotrophic lateral sclerosis patients with limb onset and the prediction of patient prognosis.

NeuronSup:基于偏见神经元抑制的深度模型去偏方法

随着深度学习的广泛应用,研究者在关注模型分类性能的同时,还需要关注模型的决策是否公平可信。存在决策偏见的深度模型会造成极大的负面影响,因此如何维持深度模型的分类正确率,同时提高模型的决策公平至关重要。目前已有工作提出了较多方法,用于改善模型的个体公平,但是这些方法仍然在去偏效果、去偏后模型可用性、去偏效率等方面存在缺陷。为此,文中分析了深度模型存在个体偏见时神经元异常激活现象,提出了一种基于偏见神经元抑制的模型去偏方法NeuronSup,具有显著降低个体偏见、对主任务性能影响小、时间复杂度低等优势。具体而言,首先根据深度模型部分神经元由于个体偏见而产生异常激活的现象提出了偏见神经元的概念。然后,利用歧视样本对查找深度模型中的偏见神经元,通过抑制偏见神经元的异常激活大幅降低深度模型的个体偏见,并且根据每个神经元的最大权重边确定主任务性能神经元,通过保持深度模型的主任务性能神经元参数不变,来减小去偏操作对深度模型分类性能造成的影响。因为NeuronSup只对深度模型中的特定神经元进行去偏操作,所以时间复杂度更低,效率更高。最后,在3个真实数据集的6种敏感属性上开展去偏实验,与5种对比算法相比,NeuronSup将个体公平指标THEMIS降低了50%以上,同时使去偏操作对深度模型分类准确率的影响降低到3%以内,验证了NeuronSup在保证深度模型分类能力的情况下降低个体偏见的有效性。

Synchronization of the minimal models of bursting neurons coupled by delayed chemical or electrical synapses

The minimal two-dimensional model of bursting neuronal dynamics is used to study the influence of time-delay on the properties of synchronization of bursting neurons. Generic properties of bursting and dependence of the stability of synchronization on the time-lag and the strength of coupling are described, and compared with the two common types of synaptical coupling, i.e., time-delayed chemical and electrical synapses.

Distinct neuronal excitability alterations of medial prefrontal cortex in early-life neglect model of rats

Object:Early-life neglect has irreversible emotional effects on the central nervous system.In this work,we aimed to elucidate distinct functional neural changes in me-dial prefrontal cortex(mPFC)of model rats.Methods:Maternal separation with early weaning was used as a rat model of early-life neglect.The excitation of glutamatergic and GABAergic neurons in rat mPFC was recorded and analyzed by whole-cell patch clamp.Results:Glutamatergic and GABAergic neurons of mPFC were distinguished by typi-cal electrophysiological properties.The excitation of mPFC glutamatergic neurons was significantly increased in male groups,while the excitation of mPFC GABAergic neurons was significant in both female and male groups,but mainly in terms of rest membrane potential and amplitude,respectively.Conclusions:Glutamatergic and GABAergic neurons in medial prefrontal cortex showed different excitability changes in a rat model of early-life neglect,which can contribute to distinct mechanisms for emotional and cognitive manifestations.

Serotonin controls axon and neuronal regeneration in the nervous system:lessons from regenerating animal models

Traumatic brain injury （TBI） is a mechanical injury to brain tissue that leads to an impairment of function and a broad spectrum of symptoms and disabilities; often, it is followed by diffuse axonal injury, which causes denaturation of the white matter and axon retraction, leaving patients with severe brain damage or even in a persistent vegetative state.

Stimulatory Effects of NMDA on Intracellular Ca^(2+) Nonlinear Kinetic Model in Rat Dorsal Root Ganglion Neurons

The present study revealed the stimulatory effects of NMDA on intracellular ca 2+ concentration in rat dorsal root ganglion (DRG) neurons. Fura 3/AM, an intracellular calcium fluorescent indicator was used to monitor the fluctuation of 〔ca 2+ 〕 i. Here we present the evidence that (1) Confocal microscopy resolved the cells of three different sizes, based on which a cell diameter distribution histogram was drawn. The fluorescence signals originated from the cells of different sizes, small size (diameter<30μm), medium size (diameter between 30 to 50μm),and large size (diameter>50μm); presumably intracellular Ca 2+ concentration was different in the cells of different sizes. (2) The time related variation of fluorescence intensity was observed. In particular, the fluorescence intensity in 0 Ca 2+ bath solution was affected by the application of high ca 2+ solution. (3) In 0 ca 2+ bath solution the intracellular Ca 2+ concentration nonlinear properties of distinct diameter cells was described. (4) A kind of SETAR model was fitted with a medium sized cell.(5)The residuals from the fitted model were tested to see whether they were plausibly Gaussian. These findings indicated that in distinct types of cells intracellular Ca 2+ concentration had different characteristics in different DRG neurons, and contributed to different functions of these neurons. Besides, the established threshold autoregressive model can share intracellular ca 2+ with the main nonlinear kinetic

A Mechanoelectrical Coupling Model of Neurons under Stretching

Introduction Neurons are situated in a microenvironment composed of various biochemical and biophysical cues,where stretching is thought to have a major impact on neurons.For instance,during a moderate traumatic brain impact,the injury region in axons exhibits significant longitudinal strain;and in a rat model of spinal cord injury,the most severe axonal injury is located in the largest strain region.Stretching may result in microstructural changes in neural tissue and further leading to abnormal electrophysiological function.Hence,it is of great importance to understand the coupled mechanoelectricalbehaviors of neurons under stretching.In spite of significant experimental efforts,the underlying mechanism remains elusive,more works are needed to provide a detailed description of the process that leads to the observed phenomena.Mathematical modeling is a powerful tool that offers a quantitative description of the underlying mechanism of an observed biological phenomenon,including mechanical and electrophysiological behaviors of neurons.Thus,we developed a mechanoelectrical coupling model of neurons under stretching in this study.Mathematical model The mathematical model consists of three submodels,i.e.,the mechanical submodel,the mechanoelectrical coupling submodel and the electrophysiological submodel.The mechanical submodel deals with the relationship between stretching and the deformation of axons,which has specially considered the plastic deformation of axons.The electrophysiological submodel characterizes the feature of neuronal action potential(AP),which is based on the classical H-H model and the cable theory.The mechanoelectrical coupling submodel links the mechanical and electrophysiological submodels through strain-induced equivalent circuit parameter alteration and ion channel injury.Besides,we have discussed a more general deformation condition,where an expanded model coupling the axonal deformation and electrophysiology alteration was explored.As the most essential parameters in an electrophysiological assessment,the amplitude of the AP,the neuronal firing frequency and the electrophysiological signal conduction velocity,which could be affected by stretching,were used as outputs of the model.Results&discussion To understand the mechanoelectrical coupling of neurons under stretching,we developed a mechanoelectrical coupling model.To verify the model,we simulated a slow stretching on an axon following the experimental study in the literature,we observed that as the strain increases,the peak AP declines faster,which is consistent with the experimental data.Moreover,the reduced AP cannot be restored to the original peak,implying that the damage is irreversible.The simulation results also predict that strain induces a more frequent neuronal firing and a faster conduction.In a realistic situation,in addition to stretching,the loading condition is very complicated,which may induce complex axonal deformation(e.g., necking and swelling along the axons).We also simulated such necking deformation impairment and observed that the AP amplitude decreases at the necking region and recovers after that,indicating a blockage of the AP;and the conduction velocity decreases with the increase in deformation degree.Conclusions In this study,we developed a mechanoelectrical coupling model of neurons under stretching with consideration of axonal plastic deformation.With the model,we found that the effect of mechanical loading on electrophysiology mainly manifests as decreased membrane AP amplitude,a more frequent neuronal firing and a faster electrophysiological signal conduction.The model predicts not only stretch-induced injury but also a more gene ral necking deformation case,which may someday be revealed in future by experiments,providing a reference for the prediction and regulation of neuronal function under mechanical loadings.

Determinants of Response Pattern of Cochlear Nucleus Neuron:a Study Based on the Model

Objective: Neurons in the cochlear nucleus show different response patterns to the short tone bursts. Because of the limitations of animal experiments, it is hard to explore the principle. Therefore, using a model to simulate CN neurons will be a feasible way. Methods: Based on the initial model mentioned in the previous study, we proposed an improved CN model in MATLAB R2012b. Results: By modifying the parameters of the model we found the interchanges among "primary-like", "chopper",and "onset" response patterns. Furthermore, we simulated the "pauser" response pattern by adding an extra input in our model. Conclusion: The results indicate that the synaptic integrations and the input modes can give rise to different characteristics of CN neurons, which eventually determine the response patterns of CN neurons.

Alleviated mucosal and neuronal damage in a rat model of Crohn's disease

AIM:To establish a rat model suitable to investigate the repetitive relapsing inflammations(RRI)characteristic to Crohn’s disease.METHODS:Colitis was induced by 2,4,6-trinitrobenzenesulfonic acid(TNBS).RRI were mimicked by repeating administrations of TNBS.Tissue samples were taken from control,once,twice and three times treated rats from the inflamed and adjacent non-inflamed colonic segments at different timepoints during the acute intestinal inflammation.The means of the ulcerated area were measured to evaluate the macroscopic mu-cosal damage.The density of myenteric neurons was determined on whole mounts by Hu C/Hu D immunohistochemistry.Heme oxygenase-1(HO-1)expression was evaluated by molecular biological techniques.RESULTS:TNBS-treated rats displayed severe colitis,but the mortality was negligible,and an increase of body weight was characteristic throughout the experimental period.The widespread loss of myenteric neurons,and marked but transient HO-1 up-regulation were demonstrated after the first TNBS administration.After repeated doses the length of the recovery time and extent of the ulcerous colonic segments were markedly decreased,and the neuronal loss was on a smaller scale and was limited to the inflamed area.HO-1 m RNA level was notably greater than after a single dose and overexpression was sustained throughout the timepoints examined.Nevertheless,the HO-1protein up-regulation after the second TNBS treatment proved to be transient.Following the third treatment HO-1 protein expression could not be detected.CONCLUSION:Experimentally provoked RRI may exert a protective preconditioning effect against the mucosal and neuronal damage.The persistent up-regulation of HO-1 m RNA expression may correlate with this.

A New Kind of Artificial Neuron Models

ＡＮｅｗＫｉｎｄｏｆＡｒｔｉｆｉｃｉａｌＮｅｕｒｏｎＭｏｄｅｌｓＦＥＮＧＹｉｎｇｉｕｎＲＡＮＱｉｗｅｎＭＡＯＷｅｎｇｅ（冯英浚）（冉启文）（毛文革）（Ｄｅｐｔ．ｏｆＭａｔｈｅｍａｔｉｃｓ，ＨａｒｂｉｎＩｎｓｔｉｔｕｔｅｏｆＴｅｃｈｎｏｌｏｇｙ，Ｈａｒ...

Using induced pluripotent stem cells derived neurons to model brain diseases

The ability to use induced pluripotent stem cells（i PSC）to model brain diseases is a powerful tool for unraveling mechanistic alterations in these disorders.Rodent models of brain diseases have spurred understanding of pathology but the concern arises that they may not recapitulate the full spectrum of neuron disruptions associated with human neuropathology.iPSC derived neurons,or other neural cell types,provide the ability to access pathology in cells derived directly from a patient＇s blood sample or skin biopsy where availability of brain tissue is limiting.Thus,utilization of iPSC to study brain diseases provides an unlimited resource for disease modelling but may also be used for drug screening for effective therapies and may potentially be used to regenerate aged or damaged cells in the future.Many brain diseases across the spectrum of neurodevelopment,neurodegenerative and neuropsychiatric are being approached by iPSC models.The goal of an iPSC based disease model is to identify a cellular phenotype that discriminates the disease-bearing cells from the control cells.In this mini-review,the importance of iPSC cell models validated for pluripotency,germline competency and function assessments is discussed.Selected examples for the variety of brain diseases that are being approached by iPSC technology to discover or establish the molecular basis of the neuropathology are discussed.

Numerical Simulation Analysis of a Mathematical Model of Circadian Pacemaker Neurons

Sim and Forger have proposed a mathematical model of circadian pacemaker neurons in the suprachiasmatic nucleus (SCN). This model, which has been formulated on the Hodgkin-Huxley mo-del, is described by a system of nonlinear ordinary differential equations. An important feature of the SCN neurons observed in electrophysiological recording is spontaneous repetitive spiking, which is reproduced using this model. In the present study, numerical simulation analysis of this model was performed to evaluate variations in two system parameters of this model: the maximal conductance of calcium current (gCa) and the maximal conductance of sodium current (gNa). Simulation results revealed the spontaneous repetitive spiking states of the model in the (gCa, gNa)-pa-rameter space.

Modeling the Spike Response for Adaptive Fuzzy Spiking Neurons with Application to a Fuzzy XOR

A spike response model(SRM)based on the spikes generator circuit(SGC)of adaptive fuzzy spiking neurons(AFSNs)is developed.The SRM is simulated in MatlabTM environment.The proposed model is applied to a configuration of a fuzzy exclusive or(fuzzy XOR)operator,as an illustrative example.A description of the comparison of AFSNs with other similar methods is given.The novel method of the AFSNs is used to determine the value of the weights or parameters of the fuzzy XOR,first with dynamic weights or self-tuning parameters that adapt continuously,then with fixed weights obtained after training,finally with fixed weights and a dynamic gain or self-tuning gain for a fine adjustment of amplitude.

Fixed Point Theorem and Fractional Differential Equations with Multiple Delays Related with Chaos Neuron Models

In this paper, we show a fixed point theorem which deduces to both of Lou’s fixed point theorem and de Pascale and de Pascale’s fixed point theorem. Moreover, our result can be applied to show the existence and uniqueness of solutions for fractional differential equations with multiple delays. Using the theorem, we discuss the fractional chaos neuron model.

Modeling and dynamics of double Hindmarsh-Rose neuron with memristor-based magnetic coupling and time delay

The firing of a neuron model is mainly affected by the following factors:the magnetic field,external forcing current,time delay,etc.In this paper,a new time-delayed electromagnetic field coupled dual Hindmarsh-Rose neuron network model is constructed.A magnetically controlled threshold memristor is improved to represent the self-connected and the coupled magnetic fields triggered by the dynamic change of neuronal membrane potential for the adjacent neurons.Numerical simulation confirms that the coupled magnetic field can activate resting neurons to generate rich firing patterns,such as spiking firings,bursting firings,and chaotic firings,and enable neurons to generate larger firing amplitudes.The study also found that the strength of magnetic coupling in the neural network also affects the number of peaks in the discharge of bursting firing.Based on the existing medical treatment background of mental illness,the effects of time lag in the coupling process against neuron firing are studied.The results confirm that the neurons can respond well to external stimuli and coupled magnetic field with appropriate time delay,and keep periodic firing under a wide range of external forcing current.

The Effect of Variations in Ionic Conductance Values on the Suppression of Repetitive Spiking in a Mathematical Model of Type-A Medial Vestibular Nucleus Neurons

A previous study has proposed a mathematical model of type-A medial vestibular nucleus neurons (mVNn). This model is described by a system of nonlinear ordinary differential equations, which is based on the Hodgkin-Huxley formalism. The type-A mVNn model contains several ionic conductances, such as the sodium conductance, calcium conductance, delayed-rectifier potassium conductance, transient potassium conductance, and calcium-dependent potassium conductance. The previous study revealed that spontaneous repetitive spiking in the type-A mVNn model can be suppressed by hyperpolarizing stimulation. However, how this suppression is affected by the ionic conductances has not been clarified in the previous study. The present study performed numerical simulation analysis of the type-A mVNn model to clarify how variations in the different ionic conductance values affect the suppression of repetitive spiking. The present study revealed that the threshold for the transition from a repetitive spiking state to a quiescent state is differentially sensitive to variations in the ionic conductances among the different types of ionic conductance.