Neural signal analysis with memristor arrays towards high-efficiency brain-machine interfaces

Single-trial detection of auditory cues from the rat brain using memristors

Implantable devices hold the potential to address conditions currently lacking effective treatments, such as drug-resistant neural impairments and prosthetic control. Medical devices need to be biologically compatible while providing enhanced performance metrics of low-power consumption, high accuracy, small size, and minimal latency to enable ongoing intervention in brain function. Here, we demonstrate a memristor-based processing system for single-trial detection of behaviorally meaningful brain signals within a timeframe that supports real-time closed-loop intervention. We record neural activity from the reward center of the brain, the ventral tegmental area, in rats trained to associate a musical tone with a reward, and we use the memristors built-in thresholding properties to detect nontrivial biomarkers in local field potentials. This approach yields consistent and accurate detection of biomarkers >98% while maintaining power consumption as low as 4.14 nanowatt per channel. The efficacy of our system's capabilities to process real-time in vivo neural data paves the way for low-power chronic neural activity monitoring and biomedical implants.

Semantic memory-based dynamic neural network using memristive ternary CIM and CAM for 2D and 3D vision

The brain is dynamic, associative, and efficient. It reconfigures by associating the inputs with past experiences, with fused memory and processing. In contrast, AI models are static, unable to associate inputs with past experiences, and run on digital computers with physically separated memory and processing. We propose a hardware-software co-design, a semantic memory-based dynamic neural network using a memristor. The network associates incoming data with the past experience stored as semantic vectors. The network and the semantic memory are physically implemented on noise-robust ternary memristor-based computing-in-memory (CIM) and content-addressable memory (CAM) circuits, respectively. We validate our co-designs, using a 40-nm memristor macro, on ResNet and PointNet++ for classifying images and three-dimensional points from the MNIST and ModelNet datasets, which achieves not only accuracy on par with software but also a 48.1 and 15.9% reduction in computational budget. Moreover, it delivers a 77.6 and 93.3% reduction in energy consumption.

Recent Progress in Artificial Neurons for Neuromodulation

Driven by the rapid advancement and practical implementation of biomaterials, fabrication technologies, and artificial intelligence, artificial neuron devices and systems have emerged as a promising technology for interpreting and transmitting neurological signals. These systems are equipped with multi-modal bio-integrable sensing capabilities, and can facilitate the benefits of neurological monitoring and modulation through accurate physiological recognition. In this article, we provide an overview of recent progress in artificial neuron technology, with a particular focus on the high-tech applications made possible by innovations in material engineering, new designs and technologies, and potential application areas. As a rapidly expanding field, these advancements have a promising potential to revolutionize personalized healthcare, human enhancement, and a wide range of other applications, making artificial neuron devices the future of brain-machine interfaces.

Approaching the Ideal Linearity in Epitaxial Crystalline-Type Memristor by Controlling Filament Growth

Brain-inspired neuromorphic computing has attracted widespread attention owing to its ability to perform parallel and energy-efficient computation. However, the synaptic weight of amorphous/polycrystalline oxide based memristor usually exhibits large nonlinear behavior with high asymmetry, which aggravates the complexity of peripheral circuit system. Controllable growth of conductive filaments is highly demanded for achieving the highly linear conductance modulation. However, the stochastic behavior of the filament growth in commonly used amorphous/polycrystalline oxide memristor makes it very challenging. Here, the epitaxially grown Hf0.5Zr0.5O2-based memristor with the linearity and symmetry approaching ideal case is reported. A layer of Cu nanoparticles is inserted into epitaxially grown Hf0.5Zr0.5O2 film to form the grain boundaries due to the breaking of the epitaxial growth. By combining with the local electric field enhancement, the growth of filament is confined in the grain boundaries due to the fact that the diffusion of oxygen vacancy in crystalline lattice is more difficult than that in the grain boundaries. Furthermore, the decimal operation and high-accuracy neural network are demonstrated by utilizing the highly linear and multi-level conductance modulation capacity. This method opens an avenue to control the filament growth for the application of resistance random access memory (RRAM) and neuromorphic computing.

IGZO/WO3-x-Heterostructured Artificial Optoelectronic Synaptic Devices Mimicking Image Segmentation and Motion Capture

Currently, artificial neural networks (ANNs) based on memristors are limited to recognizing static images of objects when simulating human visual system, preventing them from performing high-dimensional information perception, and achieving more complex biomimetic functions is subject to certain limitations. In this work, indium gallium zinc oxide (IGZO)/tungsten oxide (WO3-x)-heterostructured artificial optoelectronic synaptic devices mimicking image segmentation and motion capture exhibiting high-performance optoelectronic synaptic responses are proposed and demonstrated. Upon electrical and optical stimulations, the device shows a variety of fundamental and advanced electrical and optical synaptic plasticity. Most importantly, outstanding and repeatable linear synaptic weight changes are attained by the developed memristor. By taking advantage of the notable linear synaptic weight changes, ANNs have been constructed and successfully utilized to demonstrate two applications in the field of computer vision, including image segmentation and object tracking. The accuracy attained by the memristor-based ANNs is similar to that of the computer algorithms, while its power has been significantly reduced by 105 orders of magnitude. With successful emulations of the human brain reactions when observing objects, the demonstrated memristor and related ANNs can be effectively utilized in constructing artificial optoelectronic synaptic devices and show promising potential in emulating human visual perception.

Integration of MoS2 Memtransistor Devices and Analogue Circuits for Sensor Fusion in Autonomous Vehicle Target Localization

In contemporary autonomous driving systems relying on sensor fusion, traditional digital processors encounter challenges associated with analogue-to-digital conversion and iterative vector-matrix operations, which are encumbered by limitations in terms of response time and energy consumption. In this study, we present an analogue Kalman filter circuit based on molybdenum disulfide (MoS2) memtransistor, designed to accelerate sensor fusion for precise localization in autonomous vehicle applications. The nonvolatile memory characteristics of the memtransistor allow for the storage of a fixed Kalman gain, which eliminates the data convergence and thus accelerates the processing speeds. Additionally, the modulation of multiple conductance states by the gate terminal enables fast adaptability to diverse autonomous driving scenarios by tuning multiple Kalman filter gains. Our proposed analogue Kalman filter circuit accurately estimates the position coordinates of target vehicles by fusing sensor data from light detection and ranging (LiDAR), millimeter-wave radar (Radar), and camera, and it successfully solves real-word problems in a signal-free crossroad intersection. Notably, our system achieves a 1000-fold improvement in energy efficiency compared to that of digital circuits. This work underscores the viability of a memtransistor for achieving fast, energy-efficient real-time sensing, and continuous signal processing in advanced sensor fusion technology.

Monolayer Vacancy-Induced MXene Memory for Write-Verify-Free Programming

The fundamental logic states of 1 and 0 in Complementary Metal-Oxide-Semiconductor (CMOS) are essential for modern high-speed non-volatile solid-state memories. However, the accumulated storage signal in conventional physical components often leads to data distortion after multiple write operations. This necessitates a write-verify operation to ensure proper values within the 0/1 threshold ranges. In this work, a non-gradual switching memory with two distinct stable resistance levels is introduced, enabled by the asymmetric vertical structure of monolayer vacancy-induced oxidized Ti3C2Tx MXene for efficient carrier trapping and releasing. This non-cumulative resistance effect allows non-volatile memories to attain valid 0/1 logic levels through direct reprogramming, eliminating the need for a write-verify operation. The device exhibits superior performance characteristics, including short write/erase times (100 ns), a large switching ratio (≈3 × 104), long cyclic endurance (>104 cycles), extended retention (>4 × 106 s), and highly resistive stability (>104 continuous write operations). These findings present promising avenues for next-generation resistive memories, offering faster programming speed, exceptional write performance, and streamlined algorithms.

Neuromorphic Nanoionics for Human-Machine Interaction: From Materials to Applications

Human-machine interaction (HMI) technology has undergone significant advancements in recent years, enabling seamless communication between humans and machines. Its expansion has extended into various emerging domains, including human healthcare, machine perception, and biointerfaces, thereby magnifying the demand for advanced intelligent technologies. Neuromorphic computing, a paradigm rooted in nanoionic devices that emulate the operations and architecture of the human brain, has emerged as a powerful tool for highly efficient information processing. This paper delivers a comprehensive review of recent developments in nanoionic device-based neuromorphic computing technologies and their pivotal role in shaping the next-generation of HMI. Through a detailed examination of fundamental mechanisms and behaviors, the paper explores the ability of nanoionic memristors and ion-gated transistors to emulate the intricate functions of neurons and synapses. Crucial performance metrics, such as reliability, energy efficiency, flexibility, and biocompatibility, are rigorously evaluated. Potential applications, challenges, and opportunities of using the neuromorphic computing technologies in emerging HMI technologies, are discussed and outlooked, shedding light on the fusion of humans with machines.

Toward a Brain-Neuromorphics Interface

Brain-computer interfaces (BCIs) that enable human-machine interaction have immense potential in restoring or augmenting human capabilities. Traditional BCIs are realized based on complementary metal-oxide-semiconductor (CMOS) technologies with complex, bulky, and low biocompatible circuits, and suffer with the low energy efficiency of the von Neumann architecture. The brain-neuromorphics interface (BNI) would offer a promising solution to advance the BCI technologies and shape the interactions with machineries. Neuromorphic devices and systems are able to provide substantial computation power with extremely high energy-efficiency by implementing in-materia computing such as in situ vector-matrix multiplication (VMM) and physical reservoir computing. Recent progresses on integrating neuromorphic components with sensing and/or actuating modules, give birth to the neuromorphic afferent nerve, efferent nerve, sensorimotor loop, and so on, which has advanced the technologies for future neurorobotics by achieving sophisticated sensorimotor capabilities as the biological system. With the development on the compact artificial spiking neuron and bioelectronic interfaces, the seamless communication between a BNI and a bioentity is reasonably expectable. In this review, the upcoming BNIs are profiled by introducing the brief history of neuromorphics, reviewing the recent progresses on related areas, and discussing the future advances and challenges that lie ahead.

Aqueous chemimemristor based on proton-permeable graphene membranes

Memristive devices, electrical elements whose resistance depends on the history of applied electrical signals, are leading candidates for future data storage and neuromorphic computing. Memristive devices typically rely on solid-state technology, while aqueous memristive devices are crucial for biology-related applications such as next-generation brain-machine interfaces. Here, we report a simple graphene-based aqueous memristive device with long-term and tunable memory regulated by reversible voltage-induced interfacial acid-base equilibria enabled by selective proton permeation through the graphene. Surface-specific vibrational spectroscopy verifies that the memory of the graphene resistivity arises from the hysteretic proton permeation through the graphene, apparent from the reorganization of interfacial water at the graphene/water interface. The proton permeation alters the surface charge density on the CaF2 substrate of the graphene, affecting graphene's electron mobility, and giving rise to synapse-like resistivity dynamics. The results pave the way for developing experimentally straightforward and conceptually simple aqueous electrolyte-based neuromorphic iontronics using two-dimensional (2D) materials.

Organic iontronic memristors for artificial synapses and bionic neuromorphic computing

To tackle the current crisis of Moore's law, a sophisticated strategy entails the development of multistable memristors, bionic artificial synapses, logic circuits and brain-inspired neuromorphic computing. In comparison with conventional electronic systems, iontronic memristors offer greater potential for the manifestation of artificial intelligence and brain-machine interaction. Organic iontronic memristive materials (OIMs), which possess an organic backbone and exhibit stoichiometric ionic states, have emerged as pivotal contenders for the realization of high-performance bionic iontronic memristors. In this review, a comprehensive analysis of the progress and prospects of OIMs is presented, encompassing their inherent advantages, diverse types, synthesis methodologies, and wide-ranging applications in memristive devices. Predictably, the field of OIMs, as a rapidly developing research subject, presents an exciting opportunity for the development of highly efficient neuro-iontronic systems in areas such as in-sensor computing devices, artificial synapses, and human perception.

Soft Bioelectronics for Neuroengineering: New Horizons in the Treatment of Brain Tumor and Epilepsy

Soft bioelectronic technologies for neuroengineering have shown remarkable progress, which include novel soft material technologies and device design strategies. Such technological advances that are initiated from fundamental brain science are applied to clinical neuroscience and provided meaningful promises for significant improvement in the diagnosis efficiency and therapeutic efficacy of various brain diseases recently. System-level integration strategies in consideration of specific disease circumstances can enhance treatment effects further. Here, recent advances in soft implantable bioelectronics for neuroengineering, focusing on materials and device designs optimized for the treatment of intracranial disease environments, are reviewed. Various types of soft bioelectronics for neuroengineering are categorized and exemplified first, and then details for the sensing and stimulating device components are explained. Next, application examples of soft implantable bioelectronics to clinical neuroscience, particularly focusing on the treatment of brain tumor and epilepsy are reviewed. Finally, an ideal system of soft intracranial bioelectronics such as closed-loop-type fully-integrated systems is presented, and the remaining challenges for their clinical translation are discussed.

An ultrasmall organic synapse for neuromorphic computing

High-performance organic neuromorphic devices with miniaturized device size and computing capability are essential elements for developing brain-inspired humanoid intelligence technique. However, due to the structural inhomogeneity of most organic materials, downscaling of such devices to nanoscale and their high-density integration into compact matrices with reliable device performance remain challenging at the moment. Herein, based on the design of a semicrystalline polymer PBFCL10 with ordered structure to regulate dense and uniform formation of conductive nanofilaments, we realize an organic synapse with the smallest device dimension of 50 nm and highest integration size of 1 Kb reported thus far. The as-fabricated PBFCL10 synapses can switch between 32 conductance states linearly with a high cycle-to-cycle uniformity of 98.89% and device-to-device uniformity of 99.71%, which are the best results of organic devices. A mixed-signal neuromorphic hardware system based on the organic neuromatrix and FPGA controller is implemented to execute spiking-plasticity-related algorithm for decision-making tasks.

Bringing uncertainty quantification to the extreme-edge with memristor-based Bayesian neural networks

Safety-critical sensory applications, like medical diagnosis, demand accurate decisions from limited, noisy data. Bayesian neural networks excel at such tasks, offering predictive uncertainty assessment. However, because of their probabilistic nature, they are computationally intensive. An innovative solution utilizes memristors' inherent probabilistic nature to implement Bayesian neural networks. However, when using memristors, statistical effects follow the laws of device physics, whereas in Bayesian neural networks, those effects can take arbitrary shapes. This work overcome this difficulty by adopting a variational inference training augmented by a "technological loss", incorporating memristor physics. This technique enabled programming a Bayesian neural network on 75 crossbar arrays of 1,024 memristors, incorporating CMOS periphery for in-memory computing. The experimental neural network classified heartbeats with high accuracy, and estimated the certainty of its predictions. The results reveal orders-of-magnitude improvement in inference energy efficiency compared to a microcontroller or an embedded graphics processing unit performing the same task.

Simulating Synaptic Behaviors through Frequency Modulation in a Capacitor-Memristor Circuit

Memristors, known for their adjustable and non-volatile resistance, offer a promising avenue for emulating synapses. However, achieving pulse frequency-dependent synaptic plasticity in memristors or memristive systems necessitates further exploration. In this study, we present a novel approach to modulate the conductance of a memristor in a capacitor-memristor circuit by finely tuning the frequency of input pulses. Our experimental results demonstrate that these phenomena align with the long-term depression (LTD) and long-term potentiation (LTP) observed in synapses, which are induced by the frequency of action potentials. Additionally, we successfully implement a Hebbian-like learning mechanism in a simple circuit that connects a pair of memristors to a capacitor, resulting in observed associative memory formation and forgetting processes. Our findings highlight the potential of capacitor-memristor circuits in faithfully replicating the frequency-dependent behavior of synapses, thereby offering a valuable contribution to the development of brain-inspired neural networks.

Emerging Iontronic Neural Devices for Neuromorphic Sensory Computing

Living organisms have a very mysterious and powerful sensory computing system based on ion activity. Interestingly, studies on iontronic devices in the past few years have proposed a promising platform for simulating the sensing and computing functions of living organisms, because: 1) iontronic devices can generate, store, and transmit a variety of signals by adjusting the concentration and spatiotemporal distribution of ions, which analogs to how the brain performs intelligent functions by alternating ion flux and polarization; 2) through ionic-electronic coupling, iontronic devices can bridge the biosystem with electronics and offer profound implications for soft electronics; 3) with the diversity of ions, iontronic devices can be designed to recognize specific ions or molecules by customizing the charge selectivity, and the ionic conductivity and capacitance can be adjusted to respond to external stimuli for a variety of sensing schemes, which can be more difficult for electron-based devices. This review provides a comprehensive overview of emerging neuromorphic sensory computing by iontronic devices, highlighting representative concepts of both low-level and high-level sensory computing and introducing important material and device breakthroughs. Moreover, iontronic devices as a means of neuromorphic sensing and computing are discussed regarding the pending challenges and future directions.

Electrochemical-Memristor-Based Artificial Neurons and Synapses-Fundamentals, Applications, and Challenges

Artificial neurons and synapses are considered essential for the progress of the future brain-inspired computing, based on beyond von Neumann architectures. Here, a discussion on the common electrochemical fundamentals of biological and artificial cells is provided, focusing on their similarities with the redox-based memristive devices. The driving forces behind the functionalities and the ways to control them by an electrochemical-materials approach are presented. Factors such as the chemical symmetry of the electrodes, doping of the solid electrolyte, concentration gradients, and excess surface energy are discussed as essential to understand, predict, and design artificial neurons and synapses. A variety of two- and three-terminal memristive devices and memristive architectures are presented and their application for solving various problems is shown. The work provides an overview of the current understandings on the complex processes of neural signal generation and transmission in both biological and artificial cells and presents the state-of-the-art applications, including signal transmission between biological and artificial cells. This example is showcasing the possibility for creating bioelectronic interfaces and integrating artificial circuits in biological systems. Prospectives and challenges of the modern technology toward low-power, high-information-density circuits are highlighted.

Targeted Chemical Processing Initiating Biosome Action-Potential-Matched Artificial Synapses for the Brain-Machine Interface

A great gap still exists between artificial synapses and their biological counterparts in operation voltage or stimulation duration. Here, an artificial synaptic device based on a thin-film transistor with an operating voltage (-50-50 mV) analogous to biological action potential is developed by targeted chemical processing with the help of supercritical fluids. Chemical molecules [hexamethyldisilazane (HMDS)] are elaborately chosen and brought into the target interface to form charge receptors through supercritical processing. These charge receptors with the ability of capturing electrons mimic neurotransmitter receptors in terms of mechanism and constitute key players accounting for the synaptic behaviors. The relatively lower electrical barrier height contributes to an action-potential-matched operating voltage and considerably low power consumption (∼1 pJ/synaptic event), minimizing the divide with biological synapse for a seamless linkage to the biosystem or brain-machine interface. The stable synaptic behaviors also lead to near-ideal accuracy in pattern recognition. Moreover, this methodology that introduces chemical groups into a target interface can be viewed as a platform technology that could be adapted to other conventional devices with suitable chemical molecules to reach designed synaptic behaviors. This environmentally friendly and low-temperature processing method, which can be performed even after device fabrication, has the potential to play an important role in the future development of bionic devices.

Interfacing Biology and Electronics with Memristive Materials

Memristive technologies promise to have a large impact on modern electronics, particularly in the areas of reconfigurable computing and artificial intelligence (AI) hardware. Meanwhile, the evolution of memristive materials alongside the technological progress is opening application perspectives also in the biomedical field, particularly for implantable and lab-on-a-chip devices where advanced sensing technologies generate a large amount of data. Memristive devices are emerging as bioelectronic links merging biosensing with computation, acting as physical processors of analog signals or in the framework of advanced digital computing architectures. Recent developments in the processing of electrical neural signals, as well as on transduction and processing of chemical biomarkers of neural and endocrine functions, are reviewed. It is concluded with a critical perspective on the future applicability of memristive devices as pivotal building blocks in bio-AI fusion concepts and bionic schemes.

Influence of junction resistance on spatiotemporal dynamics and reservoir computing performance arising from an SWNT/POM 3D network formed via a scaffold template technique

For scientists in numerous fields, creating a physical device that can function like the human brain is an aspiration. It is believed that we may achieve brain-like spatiotemporal information processing by fabricating an in materio reservoir computing (RC) device because of a complex random network topology with nonlinear dynamics. One of the significant drawbacks of a two-dimensional physical reservoir system is the difficulty in controlling the network density. This work reports the use of a 3D porous template as a scaffold to fabricate a three-dimensional network of a single-walled carbon nanotube polyoxometalate nanocomposite. Although the three-dimensional system exhibits better nonlinear dynamics and spatiotemporal dynamics, and higher harmonics generation than a two-dimensional system, the results suggest a correlation between a higher number of resistive junctions and reservoir performance. We show that by increasing the spatial dimension of the device, the memory capacity improves, while the scale-free network exponent (γ) remains nearly unchanged. The three-dimensional device also displays improved performance in the well-known RC benchmark task of waveform generation. This study demonstrates the impact of an additional spatial dimension, network distribution and network density on in materio RC device performance and tries to shed some light on the reason behind such behavior.

Cellular automata imbedded memristor-based recirculated logic in-memory computing

Memristor-based circuits offer low hardware costs and in-memory computing, but full-memristive circuit integration for different algorithm remains limited. Cellular automata (CA) has been noticed for its well-known parallel, bio-inspired, computational characteristics. Running CA on conventional chips suffers from low parallelism and high hardware costs. Establishing dedicated hardware for CA remains elusive. We propose a recirculated logic operation scheme (RLOS) using memristive hardware and 2D transistors for CA evolution, significantly reducing hardware complexity. RLOS's versatility supports multiple CA algorithms on a single circuit, including elementary CA rules and more complex majority classification and edge detection algorithms. Results demonstrate up to a 79-fold reduction in hardware costs compared to FPGA-based approaches. RLOS-based reservoir computing is proposed for edge computing development, boasting the lowest hardware cost (6 components/per cell) among existing implementations. This work advances efficient, low-cost CA hardware and encourages edge computing hardware exploration.

Energy-efficient high-fidelity image reconstruction with memristor arrays for medical diagnosis

Medical imaging is an important tool for accurate medical diagnosis, while state-of-the-art image reconstruction algorithms raise critical challenges in massive data processing for high-speed and high-quality imaging. Here, we present a memristive image reconstructor (MIR) to greatly accelerate image reconstruction with discrete Fourier transformation (DFT) by computing-in-memory (CIM) with memristor arrays. A high-accuracy quasi-analogue mapping (QAM) method and generic complex matrix transfer (CMT) scheme was proposed to improve the mapping precision and transfer efficiency, respectively. High-fidelity magnetic resonance imaging (MRI) and computed tomography (CT) image reconstructions were demonstrated, achieving software-equivalent qualities and DICE scores after segmentation with nnU-Net algorithm. Remarkably, our MIR exhibited 153× and 79× improvements in energy efficiency and normalized image reconstruction speed, respectively, compared to graphics processing unit (GPU). This work demonstrates MIR as a promising high-fidelity image reconstruction platform for future medical diagnosis, and also largely extends the application of memristor-based CIM beyond artificial neural networks.

Echo state graph neural networks with analogue random resistive memory arrays

Recent years have witnessed a surge of interest in learning representations of graph-structured data, with applications from social networks to drug discovery. However, graph neural networks, the machine learning models for handling graph-structured data, face significant challenges when running on conventional digital hardware, including the slowdown of Moore’s law due to transistor scaling limits and the von Neumann bottleneck incurred by physically separated memory and processing units, as well as a high training cost. Here we present a hardware–software co-design to address these challenges, by designing an echo state graph neural network based on random resistive memory arrays, which are built from low-cost, nanoscale and stackable resistors for efficient in-memory computing. This approach leverages the intrinsic stochasticity of dielectric breakdown in resistive switching to implement random projections in hardware for an echo state network that effectively minimizes the training complexity thanks to its fixed and random weights. The system demonstrates state-of-the-art performance on both graph classification using the MUTAG and COLLAB datasets and node classification using the CORA dataset, achieving 2.16×, 35.42× and 40.37× improvements in energy efficiency for a projected random resistive memory-based hybrid analogue–digital system over a state-of-the-art graphics processing unit and 99.35%, 99.99% and 91.40% reductions of backward pass complexity compared with conventional graph learning. The results point to a promising direction for next-generation artificial intelligence systems for graph learning.

Large-Scale Integrated Flexible Tactile Sensor Array for Sensitive Smart Robotic Touch

In the long pursuit of smart robotics, it has been envisioned to empower robots with human-like senses, especially vision and touch. While tremendous progress has been made in image sensors and computer vision over the past decades, tactile sense abilities are lagging behind due to the lack of large-scale flexible tactile sensor array with high sensitivity, high spatial resolution, and fast response. In this work, we have demonstrated a 64 × 64 flexible tactile sensor array with a record-high spatial resolution of 0.9 mm (equivalently 28.2 pixels per inch) by integrating a high-performance piezoresistive film (PRF) with a large-area active matrix of carbon nanotube thin-film transistors. PRF with self-formed microstructures exhibited high pressure-sensitivity of ∼385 kPa^-1 for multi-walled carbon nanotubes concentration of 6%, while the 14% one exhibited fast response time of ∼3 ms, good linearity, broad detection range beyond 1400 kPa, and excellent cyclability over 3000 cycles. Using this fully integrated tactile sensor array, the footprint maps of an artificial honeybee were clearly identified. Furthermore, we hardware-implemented a smart tactile system by integrating the PRF-based sensor array with a memristor-based computing-in-memory chip to record and recognize handwritten digits and Chinese calligraphy, achieving high classification accuracies of 98.8% and 97.3% in hardware, respectively. The integration of sensor networks with deep learning hardware may enable edge or near-sensor computing with significantly reduced power consumption and latency. Our work could empower the building of large-scale intelligent sensor networks for next-generation smart robotics.

A 13 µW Analog Front-End with RRAM-Based Lowpass FIR Filter for EEG Signal Detection

This brief presents an analog front-end (AFE) for the detection of electroencephalogram (EEG) signals. The AFE is composed of four sections, chopper-stabilized amplifiers, ripple suppression circuit, RRAM-based lowpass FIR filter, and 8-bit SAR ADC. This is the first time that an RRAM-based lowpass FIR filter has been introduced in an EEG AFE, where the bio-plausible characteristics of RRAM are utilized to analyze signals in the analog domain with high efficiency. The preamp uses the symmetrical OTA structure, reducing power consumption while meeting gain requirements. The ripple suppression circuit greatly improves noise characteristics and offset voltage. The RRAM-based low-pass filter achieves a 40 Hz cutoff frequency, which is suitable for the analysis of EEG signals. The SAR ADC adopts a segmented capacitor structure, effectively reducing the capacitor switching power consumption. The chip prototype is designed in 40 nm CMOS technology. The overall power consumption is approximately 13 µW, achieving ultra-low-power operation.

Seizure Detection and Prediction by Parallel Memristive Convolutional Neural Networks

During the past two decades, epileptic seizure detection and prediction algorithms have evolved rapidly. However, despite significant performance improvements, their hardware implementation using conventional technologies, such as Complementary Metal-Oxide-Semiconductor (CMOS), in power and area-constrained settings remains a challenging task; especially when many recording channels are used. In this paper, we propose a novel low-latency parallel Convolutional Neural Network (CNN) architecture that has between 2-2,800x fewer network parameters compared to State-Of-The-Art (SOTA) CNN architectures and achieves 5-fold cross validation accuracy of 99.84% for epileptic seizure detection, and 99.01% and 97.54% for epileptic seizure prediction, when evaluated using the University of Bonn Electroencephalogram (EEG), CHB-MIT and SWEC-ETHZ seizure datasets, respectively. We subsequently implement our network onto analog crossbar arrays comprising Resistive Random-Access Memory (RRAM) devices, and provide a comprehensive benchmark by simulating, laying out, and determining hardware requirements of the CNN component of our system. We parallelize the execution of convolution layer kernels on separate analog crossbars to enable 2 orders of magnitude reduction in latency compared to SOTA hybrid Memristive-CMOS Deep Learning (DL) accelerators. Furthermore, we investigate the effects of non-idealities on our system and investigate Quantization Aware Training (QAT) to mitigate the performance degradation due to low Analog-to-Digital Converter (ADC)/Digital-to-Analog Converter (DAC) resolution. Finally, we propose a stuck weight offsetting methodology to mitigate performance degradation due to stuck [Formula: see text] memristor weights, recovering up to 32% accuracy, without requiring retraining. The CNN component of our platform is estimated to consume approximately 2.791 W of power while occupying an area of 31.255 mm² in a 22 nm FDSOI CMOS process.

A CMOS-memristor hybrid system for implementing stochastic binary spike timing-dependent plasticity

This paper describes a fully experimental hybrid system in which a [Formula: see text] memristive crossbar spiking neural network (SNN) was assembled using custom high-resistance state memristors with analogue CMOS neurons fabricated in 180 nm CMOS technology. The custom memristors used NMOS selector transistors, made available on a second 180 nm CMOS chip. One drawback is that memristors operate with currents in the micro-amperes range, while analogue CMOS neurons may need to operate with currents in the pico-amperes range. One possible solution was to use a compact circuit to scale the memristor-domain currents down to the analogue CMOS neuron domain currents by at least 5-6 orders of magnitude. Here, we proposed using an on-chip compact current splitter circuit based on MOS ladders to aggressively attenuate the currents by over 5 orders of magnitude. This circuit was added before each neuron. This paper describes the proper experimental operation of an SNN circuit using a [Formula: see text] 1T1R synaptic crossbar together with four post-synaptic CMOS circuits, each with a 5-decade current attenuator and an integrate-and-fire neuron. It also demonstrates one-shot winner-takes-all training and stochastic binary spike-timing-dependent-plasticity learning using this small system. This article is part of the theme issue 'Advanced neurotechnologies: translating innovation for health and well-being'.

Memristor-Based Neuromodulation Device for Real-Time Monitoring and Adaptive Control of Neuronal Populations

Neurons are specialized cells for information transmission and information processing. In fact, many neurologic disorders are directly linked not to cellular viability/homeostasis issues but rather to specific anomalies in electrical activity dynamics. Consequently, therapeutic strategies based on the direct modulation of neuronal electrical activity have been producing remarkable results, with successful examples ranging from cochlear implants to deep brain stimulation. Developments in these implantable devices are hindered, however, by important challenges such as power requirements, size factor, signal transduction, and adaptability/computational capabilities. Memristors, neuromorphic nanoscale electronic components able to emulate natural synapses, provide unique properties to address these constraints, and their use in neuroprosthetic devices is being actively explored. Here, we demonstrate, for the first time, the use of memristive devices in a clinically relevant setting where communication between two neuronal populations is conditioned to specific activity patterns in the source population. In our approach, the memristor device performs a pattern detection computation and acts as an artificial synapse capable of reversible short-term plasticity. Using in vitro hippocampal neuronal cultures, we show real-time adaptive control with a high degree of reproducibility using our monitor-compute-actuate paradigm. We envision very similar systems being used for the automatic detection and suppression of seizures in epileptic patients.

Memristor-based analogue computing for brain-inspired sound localization with in situ training

The human nervous system senses the physical world in an analogue but efficient way. As a crucial ability of the human brain, sound localization is a representative analogue computing task and often employed in virtual auditory systems. Different from well-demonstrated classification applications, all output neurons in localization tasks contribute to the predicted direction, introducing much higher challenges for hardware demonstration with memristor arrays. In this work, with the proposed multi-threshold-update scheme, we experimentally demonstrate the in-situ learning ability of the sound localization function in a 1K analogue memristor array. The experimental and evaluation results reveal that the scheme improves the training accuracy by ∼45.7% compared to the existing method and reduces the energy consumption by ∼184× relative to the previous work. This work represents a significant advance towards memristor-based auditory localization system with low energy consumption and high performance.

Perlecan, A Multi-Functional, Cell-Instructive, Matrix-Stabilizing Proteoglycan With Roles in Tissue Development Has Relevance to Connective Tissue Repair and Regeneration

This review highlights the multifunctional properties of perlecan (HSPG2) and its potential roles in repair biology. Perlecan is ubiquitous, occurring in vascular, cartilaginous, adipose, lymphoreticular, bone and bone marrow stroma and in neural tissues. Perlecan has roles in angiogenesis, tissue development and extracellular matrix stabilization in mature weight bearing and tensional tissues. Perlecan contributes to mechanosensory properties in cartilage through pericellular interactions with fibrillin-1, type IV, V, VI and XI collagen and elastin. Perlecan domain I - FGF, PDGF, VEGF and BMP interactions promote embryonic cellular proliferation, differentiation, and tissue development. Perlecan domain II, an LDLR-like domain interacts with lipids, Wnt and Hedgehog morphogens. Perlecan domain III binds FGF-7 and 18 and has roles in the secretion of perlecan. Perlecan domain IV, an immunoglobulin repeat domain, has cell attachment and matrix stabilizing properties. Perlecan domain V promotes tissue repair through interactions with VEGF, VEGF-R2 and α2β1 integrin. Perlecan domain-V LG1-LG2 and LG3 fragments antagonize these interactions. Perlecan domain V promotes reconstitution of the blood brain barrier damaged by ischemic stroke and is neurogenic and neuroprotective. Perlecan-VEGF-VEGFR2, perlecan-FGF-2 and perlecan-PDGF interactions promote angiogenesis and wound healing. Perlecan domain I, III and V interactions with platelet factor-4 and megakaryocyte and platelet inhibitory receptor promote adhesion of cells to implants and scaffolds in vascular repair. Perlecan localizes acetylcholinesterase in the neuromuscular junction and is of functional significance in neuromuscular control. Perlecan mutation leads to Schwartz-Jampel Syndrome, functional impairment of the biomechanical properties of the intervertebral disc, variable levels of chondroplasia and myotonia. A greater understanding of the functional working of the neuromuscular junction may be insightful in therapeutic approaches in the treatment of neuromuscular disorders. Tissue engineering of salivary glands has been undertaken using bioactive peptides (TWSKV) derived from perlecan domain IV. Perlecan TWSKV peptide induces differentiation of salivary gland cells into self-assembling acini-like structures that express salivary gland biomarkers and secrete α-amylase. Perlecan also promotes chondroprogenitor stem cell maturation and development of pluripotent migratory stem cell lineages, which participate in diarthrodial joint formation, and early cartilage development. Recent studies have also shown that perlecan is prominently expressed during repair of adult human articular cartilage. Perlecan also has roles in endochondral ossification and bone development. Perlecan domain I hydrogels been used in tissue engineering to establish heparin binding growth factor gradients that promote cell migration and cartilage repair. Perlecan domain I collagen I fibril scaffolds have also been used as an FGF-2 delivery system for tissue repair. With the availability of recombinant perlecan domains, the development of other tissue repair strategies should emerge in the near future. Perlecan co-localization with vascular elastin in the intima, acts as a blood shear-flow endothelial sensor that regulates blood volume and pressure and has a similar role to perlecan in canalicular fluid, regulating bone development and remodeling. This complements perlecan's roles in growth plate cartilage and in endochondral ossification to form the appendicular and axial skeleton. Perlecan is thus a ubiquitous, multifunctional, and pleomorphic molecule of considerable biological importance. A greater understanding of its diverse biological roles and functional repertoires during tissue development, growth and disease will yield valuable insights into how this impressive proteoglycan could be utilized successfully in repair biology.

Rotating neurons for all-analog implementation of cyclic reservoir computing

Hardware implementation in resource-efficient reservoir computing is of great interest for neuromorphic engineering. Recently, various devices have been explored to implement hardware-based reservoirs. However, most studies were mainly focused on the reservoir layer, whereas an end-to-end reservoir architecture has yet to be developed. Here, we propose a versatile method for implementing cyclic reservoirs using rotating elements integrated with signal-driven dynamic neurons, whose equivalence to standard cyclic reservoir algorithm is mathematically proven. Simulations show that the rotating neuron reservoir achieves record-low errors in a nonlinear system approximation benchmark. Furthermore, a hardware prototype was developed for near-sensor computing, chaotic time-series prediction and handwriting classification. By integrating a memristor array as a fully-connected output layer, the all-analog reservoir computing system achieves 94.0% accuracy, while simulation shows >1000× lower system-level power than prior works. Therefore, our work demonstrates an elegant rotation-based architecture that explores hardware physics as computational resources for high-performance reservoir computing.

Reservoir computing has demonstrated high-level performance, however efficient hardware implementations demand an architecture with minimum system complexity. The authors propose a rotating neuron-based architecture for physically implementing all-analog resource efficient reservoir computing system.

Enhancing the Recognition Task Performance of MEMS Resonator-Based Reservoir Computing System via Nonlinearity Tuning

Reservoir computing (RC) is a potential neuromorphic paradigm for physically realizing artificial intelligence systems in the Internet of Things society, owing to its well-known low training cost and compatibility with nonlinear devices. Micro-electro-mechanical system (MEMS) resonators exhibiting rich nonlinear dynamics and fading behaviors are promising candidates for high-performance hardware RC. Previously, we presented a non-delay-based RC using one single micromechanical resonator with hybrid nonlinear dynamics. Here, we innovatively introduce a nonlinear tuning strategy to analyze the computing properties (the processing speed and recognition accuracy) of the presented RC. Meanwhile, we numerically and experimentally analyze the influence of the hybrid nonlinear dynamics using the image classification task. Specifically, we study the transient nonlinear saturation phenomenon by fitting quality factors under different vacuums, as well as searching the optimal operating point (the edge of chaos) by the static bifurcation analysis and dynamic vibration numerical models of the Duffing nonlinearity. Our results in the optimal operation conditions experimentally achieved a high classification accuracy of (93 ± 1)% and several times faster than previous work on the handwritten digits recognition benchmark, profit from the perfect high signal-to-noise ratios (quality factor) and the nonlinearity of the dynamical variables.

Application of mathematical morphology operation with memristor-based computation-in-memory architecture for detecting manufacturing defects

Mathematical morphology operations are widely used in image processing such as defect analysis in semiconductor manufacturing and medical image analysis. These data-intensive applications have high requirements during hardware implementation that are challenging for conventional hardware platforms such as central processing units (CPUs) and graphics processing units (GPUs). Computation-in-memory (CIM) provides a possible solution for highly efficient morphology operations. In this study, we demonstrate the application of morphology operation with a novel memristor-based auto-detection architecture and demonstrate non-neuromorphic computation on a multi-array-based memristor system. Pixel-by-pixel logic computations with low parallelism are converted to parallel operations using memristors. Moreover, hardware-implemented computer-integrated manufacturing was used to experimentally demonstrate typical defect detection tasks in integrated circuit (IC) manufacturing and medical image analysis. In addition, we developed a new implementation scheme employing a four-layer network to realize small-object detection with high parallelism. The system benchmark based on the hardware measurement results showed significant improvement in the energy efficiency by approximately 358 times and 32 times more than when a CPU and GPU were employed, respectively, exhibiting the advantage of the proposed memristor-based morphology operation.

Reservoir computing with biocompatible organic electrochemical networks for brain-inspired biosignal classification

Early detection of malign patterns in patients' biological signals can save millions of lives. Despite the steady improvement of artificial intelligence-based techniques, the practical clinical application of these methods is mostly constrained to an offline evaluation of the patients' data. Previous studies have identified organic electrochemical devices as ideal candidates for biosignal monitoring. However, their use for pattern recognition in real time was never demonstrated. Here, we produce and characterize brain-inspired networks composed of organic electrochemical transistors and use them for time-series predictions and classification tasks using the reservoir computing approach. To show their potential use for biofluid monitoring and biosignal analysis, we classify four classes of arrhythmic heartbeats with an accuracy of 88%. The results of this study introduce a previously unexplored paradigm for biocompatible computational platforms and may enable development of ultralow-power consumption hardware-based artificial neural networks capable of interacting with body fluids and biological tissues.

Hybrid oxide brain-inspired neuromorphic devices for hardware implementation of artificial intelligence

The state-of-the-art artificial intelligence technologies mainly rely on deep learning algorithms based on conventional computers with classical von Neumann computing architectures, where the memory and processing units are separated resulting in an enormous amount of energy and time consumed in the data transfer process. Inspired by the human brain acting like an ultra-highly efficient biological computer, neuromorphic computing is proposed as a technology for hardware implementation of artificial intelligence. Artificial synapses are the main component of a neuromorphic computing architecture. Memristors are considered to be a relatively ideal candidate for artificial synapse applications due to their high scalability and low power consumption. Oxides are most widely used in memristors due to the ease of fabrication and high compatibility with complementary metal-oxide-semiconductor processes. However, oxide memristors suffer from unsatisfactory stability and reliability. Oxide-based hybrid structures can effectively improve the device stability and reliability, therefore providing a promising prospect for the application of oxide memristors to neuromorphic computing. This work reviews the recent advances in the development of hybrid oxide memristive synapses. The discussion is organized according to the blending schemes as well as the working mechanisms of hybrid oxide memristors.

In-sensor reservoir computing for language learning via two-dimensional memristors

The dynamic processing of optoelectronic signals carrying temporal and sequential information is critical to various machine learning applications including language processing and computer vision. Despite extensive efforts to emulate the visual cortex of human brain, large energy/time overhead and extra hardware costs are incurred by the physically separated sensing, memory, and processing units. The challenge is further intensified by the tedious training of conventional recurrent neural networks for edge deployment. Here, we report in-sensor reservoir computing for language learning. High dimensionality, nonlinearity, and fading memory for the in-sensor reservoir were achieved via two-dimensional memristors based on tin sulfide (SnS), uniquely having dual-type defect states associated with Sn and S vacancies. Our in-sensor reservoir computing demonstrates an accuracy of 91% to classify short sentences of language, thus shedding light on a low training cost and the real-time solution for processing temporal and sequential signals for machine learning applications at the edge.

Research on Pt/NiOx/WO3-x:Ti/W Multijunction Memristors with Synaptic Learning and Memory Functions

Artificial synapses based on biological synapses represent a new idea in the field of artificial intelligence with future applications. Current two-terminal RRAM devices have developed tremendously due to the adjustable synaptic plasticity of artificial synapses. However, these devices still have some problems, such as current leakage and poor durability. Here, we demonstrate a Pt/NiO_x/WO_3-x:Ti/W memristor with a pn-type heterojunction and two metal-semiconductor contacts, which exhibits good rectification. Due to the change in the internal potential barrier, the devices possess multiconductance states under different pulse modulations and memory characteristics, similar to synapses. The rectification characteristics of the device exhibit stable enhancement and suppression behavior. Each device in the 10 × 10 cross array we constructed can be written correctly, which verifies that leakage current does not appear in the device. The structure proposed in this work has great significance for the integration of large-scale memristor cross arrays.