A Low-Power Impedance-to-Frequency Converter for Frequency-Multiplexed Wearable Sensors.

We propose a low-power impedance-to-frequency (I-to-F) converter for wearable transducers that change both its resistance and capacitance in response to mechanical deformation or changes in ambient pressure. At the core of the proposed I-to-F converter is a fixed-point circuit comprising of a voltage-controlled relaxation oscillator and a proportional-to-temperature (PTAT) current reference that locks the oscillation frequency according to the impedance of the transducer. Using both analytical and measurement results we show that the operation of the proposed I-to-F converter is well matched to a specific class of sponge mechanical transducer where the system can achieve higher sensitivity when compared to a simple resistance measurement techniques. Furthermore, the oscillation frequency of the converter can be programmed to ensure that multiple transducer and I-to-F converters can communicate simultaneously over a shared channel (physical wire or virtual wireless channel) using frequency-division multiplexing. Measured results from proof-of-concept prototypes show an impedance sensitivity of 19.66 Hz/ Ω at 1.1 kΩ load impedance magnitude and a current consumption of [Formula: see text]. As a demonstration we show the application of the I-to-F converter for human gesture recognition and for radial pulse sensing.

Augmenting insect olfaction performance through nano-neuromodulation.

Biological olfactory systems are highly sensitive and selective, often outperforming engineered chemical sensors in highly complex and dynamic environments. As a result, there is much interest in using biological systems to build sensors. However, approaches to read-out information from biological systems, especially neural signals, tend to be suboptimal due to the number of electrodes that can be used and where these can be placed. Here we aim to overcome this suboptimality in neural information read-out by using a nano-enabled neuromodulation strategy to augment insect olfaction-based chemical sensors. By harnessing the photothermal properties of nanostructures and releasing a select neuromodulator on demand, we show that the odour-evoked response from the interrogated regions of the insect olfactory system can not only be enhanced but can also improve odour identification.

A Portable and a Scalable Multi-Channel Wireless Recording System for Wearable Electromyometrial Imaging.

Electromyometrial imaging (EMMI) technology has emerged as one of the promising technology that can be used for non-invasive pregnancy risk stratification and for preventing complications due to pre-term birth. Current EMMI systems are bulky and require a tethered connection to desktop instrumentation, as a result, the system cannot be used in non-clinical and ambulatory settings. In this article, we propose an approach for designing a scalable, portable wireless EMMI recording system that can be used for in-home and remote monitoring. The wearable system uses a non-equilibrium differential electrode multiplexing approach to enhance signal acquisition bandwidth and to reduce the artifacts due to electrode drifts, amplifier 1/f noise, and bio-potential amplifier saturation. A combination of active shielding, a passive filter network, and a high-end instrumentation amplifier ensures sufficient input dynamic range ([Formula: see text]) such that the system can simultaneously acquire different bio-potential signals like maternal electrocardiogram (ECG) in addition to the EMMI electromyogram (EMG) signals. We show that the switching artifacts and the channel cross-talk introduced due to non-equilibrium sampling can be reduced using a compensation technique. This enables the system to be potentially scaled to a large number of channels without significantly increasing the system power dissipation. We demonstrate the feasibility of the proposed approach in a clinical setting using an 8-channel battery-powered prototype which dissipates less than 8 µW per channel for a signal bandwidth of 1 KHz.

Stretchable Sponge Electrodes for Long-Term and Motion-Artifact-Tolerant Recording of High-Quality Electrophysiologic Signals.

Soft electronic devices and sensors have shown great potential for wearable and ambulatory electrophysiologic signal monitoring applications due to their light weight, ability to conform to human skin, and improved wearing comfort, and they may replace the conventional rigid electrodes and bulky recording devices widely used nowadays in clinical settings. Herein, we report an elastomeric sponge electrode that offers greatly reduced electrode-skin contact impedance, an improved signal-to-noise ratio (SNR), and is ideally suited for long-term and motion-artifact-tolerant recording of high-quality biopotential signals. The sponge electrode utilizes a porous polydimethylsiloxane sponge made from a sacrificial template of sugar cubes, and it is subsequently coated with a poly(3,4-ethylenedioxythiophene) polystyrenesulfonate (PEDOT:PSS) conductive polymer using a simple dip-coating process. The sponge electrode contains numerous micropores that greatly increase the skin-electrode contact area and help lower the contact impedance by a factor of 5.25 or 6.7 compared to planar PEDOT:PSS electrodes or gold-standard Ag/AgCl electrodes, respectively. The lowering of contact impedance resulted in high-quality electrocardiogram (ECG) and electromyogram (EMG) recordings with improved SNR. Furthermore, the porous structure also allows the sponge electrode to hold significantly more conductive gel compared to conventional planar electrodes, thereby allowing them to be used for long recording sessions with minimal signal degradation. The conductive gel absorbed into the micropores also serves as a buffer layer to help mitigate motion artifacts, which is crucial for recording on ambulatory patients. Lastly, to demonstrate its feasibility and potential for clinical usage, we have shown that the sponge electrode can be used to monitor uterine contraction activities from a patient in labor. With its low-cost fabrication, softness, and ability to record high SNR biopotential signals, the sponge electrode is a promising platform for long-term wearable health monitoring applications.

An adaptive synaptic array using Fowler-Nordheim dynamic analog memory.

In this paper we present an adaptive synaptic array that can be used to improve the energy-efficiency of training machine learning (ML) systems. The synaptic array comprises of an ensemble of analog memory elements, each of which is a micro-scale dynamical system in its own right, storing information in its temporal state trajectory. The state trajectories are then modulated by a system level learning algorithm such that the ensemble trajectory is guided towards the optimal solution. We show that the extrinsic energy required for state trajectory modulation can be matched to the dynamics of neural network learning which leads to a significant reduction in energy-dissipated for memory updates during ML training. Thus, the proposed synapse array could have significant implications in addressing the energy-efficiency imbalance between the training and the inference phases observed in artificial intelligence (AI) systems.

A Soft Sponge Sensor for Multimodal Sensing and Distinguishing of Pressure, Strain, and Temperature.

Soft wearable sensors are essential components for applications such as motion tracking, human-machine interface, and soft robots. However, most of the reported sensors are either specifically designed to target an individual stimulus or capable of responding to multiple stimuli (e.g., pressure and strain) but without the necessary selectivity to distinguish those stimuli. Here we report an elastomeric sponge-based sensor that can respond to and distinguish three different kinds of stimuli: pressure, strain, and temperature. The sensor utilizes a porous polydimethylsiloxane (PDMS) sponge fabricated from a sugar cube sacrificial template, which was subsequently coated with a poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) conductive polymer through a low-cost dip-coating process. Responses to different types of stimuli can be distinguished by simultaneously recording resistance and capacitance changes. Because pressure, tensile strain, and temperature change result in different trends in resistance and capacitance change, those stimuli can be clearly distinguished from each other by simultaneously measuring the resistance and capacitance of the sensor. We have also studied the effect of the pore size on the sensor performance and have found that the sponge sensor with smaller pores generally offers greater resistance change and better sensitivity. As a proof-of-concept, we have demonstrated the use of the porous sponge sensor on an artificial hand for object detection, gesture recognition, and temperature sensing applications.

In Situ Grown Gold Nanoisland-Based Chemiresistive Electronic Nose for Sniffing Distinct Odor Fingerprints.

Chemiresistors based on metal-insulator-metal structures are attractive transducers for rapid tracing of a wide repertoire of (bio)chemical species in the vapor phase. However, current fabrication techniques suffer greatly from sensor-to-sensor variability, limiting their reproducible and reliable application in real-world settings. We demonstrate a novel, facile, and ubiquitously applicable strategy for fabricating highly reliable and reproducible organothiol-functionalized gold nanoisland-based chemiresistors. The novel fabrication technique involves iterative in situ seeding, growth, and surface functionalization of gold nanoislands on an interdigitated electrode, which in turn generates a multi-layered densely packed continuous gold nanoisland film. The chemiresistors fabricated using the proposed strategy exhibited high sensor-to-sensor reproducibility owing to the controlled iterative seeding and growth-based fabrication technique, long-term stability, and specificity for detection and identification of a wide variety of volatile organic compounds. Upon exposure to a specific odor, the chemiresistor ensemble comprised nine different chemical functionalities and produced a unique and discernable odor fingerprint that is reproducible for at least up to 90 days. Integrating these odor fingerprints with a simple linear classifier was found to be sufficient for discriminating between all six odors used in this study. We believe that the fabrication strategy presented here, which is agnostic to chemical functionality, enables fabrication of highly reliable and reproducible sensing elements, and thereby an adaptable electronic nose for a wide variety of real-world gas sensing applications.

Studying the Feasibility of Postoperative Monitoring of Spinal Fusion Progress Using a Self-Powered Fowler-Nordheim Sensor-Data-Logger.

OBJECTIVE: This study investigates the feasibility of using a new self-powered sensing and data logging system for postoperative monitoring of spinal fusion progress. The proposed diagnostic technology directly couples a piezoelectric transducer signal into a Fowler-Nordheim (FN) quantum tunneling-based synchronized dynamical system to record the mechanical usage of spinal fixation devices. The operation of the proposed implantable FN sensor-data-logger is completely self-powered by harvesting the energy from the micro-motion of the spine during the course of fusion. Bench-top testing is performed using corpectomy models to evaluate the performance of the proposed monitoring system. In order to simulate the spinal fusion process, different materials with gradually increasing elastic modulus are used to fill the intervertebral space gap. Besides, finite element models are developed to analyze the strains induced on the spinal rods during the applied cyclic loading. Data measured from the benchtop experiment is processed using an FN sensor-data-logger model to obtain time-evolution curves representing each spinal fusion state. This feasibility study shows that the obtained curves are viable tools to differentiate between conditions of osseous union and assess the effective fusion period.

On-device synaptic memory consolidation using Fowler-Nordheim quantum-tunneling.

Introduction: For artificial synapses whose strengths are assumed to be bounded and can only be updated with finite precision, achieving optimal memory consolidation using primitives from classical physics leads to synaptic models that are too complex to be scaled in-silico. Here we report that a relatively simple differential device that operates using the physics of Fowler-Nordheim (FN) quantum-mechanical tunneling can achieve tunable memory consolidation characteristics with different plasticity-stability trade-offs. Methods: A prototype FN-synapse array was fabricated in a standard silicon process and was used to verify the optimal memory consolidation characteristics and used for estimating the parameters of an FN-synapse analytical model. The analytical model was then used for large-scale memory consolidation and continual learning experiments. Results: We show that compared to other physical implementations of synapses for memory consolidation, the operation of the FN-synapse is near-optimal in terms of the synaptic lifetime and the consolidation properties. We also demonstrate that a network comprising FN-synapses outperforms a comparable elastic weight consolidation (EWC) network for some benchmark continual learning tasks. Discussions: With an energy footprint of femtojoules per synaptic update, we believe that the proposed FN-synapse provides an ultra-energy-efficient approach for implementing both synaptic memory consolidation and continual learning on a physical device.

A Sparsity-Driven Backpropagation-Less Learning Framework Using Populations of Spiking Growth Transform Neurons.

Growth-transform (GT) neurons and their population models allow for independent control over the spiking statistics and the transient population dynamics while optimizing a physically plausible distributed energy functional involving continuous-valued neural variables. In this paper we describe a backpropagation-less learning approach to train a network of spiking GT neurons by enforcing sparsity constraints on the overall network spiking activity. The key features of the model and the proposed learning framework are: (a) spike responses are generated as a result of constraint violation and hence can be viewed as Lagrangian parameters; (b) the optimal parameters for a given task can be learned using neurally relevant local learning rules and in an online manner; (c) the network optimizes itself to encode the solution with as few spikes as possible (sparsity); (d) the network optimizes itself to operate at a solution with the maximum dynamic range and away from saturation; and (e) the framework is flexible enough to incorporate additional structural and connectivity constraints on the network. As a result, the proposed formulation is attractive for designing neuromorphic tinyML systems that are constrained in energy, resources, and network structure. In this paper, we show how the approach could be used for unsupervised and supervised learning such that minimizing a training error is equivalent to minimizing the overall spiking activity across the network. We then build on this framework to implement three different multi-layer spiking network architectures with progressively increasing flexibility in training and consequently, sparsity. We demonstrate the applicability of the proposed algorithm for resource-efficient learning using a publicly available machine olfaction dataset with unique challenges like sensor drift and a wide range of stimulus concentrations. In all of these case studies we show that a GT network trained using the proposed learning approach is able to minimize the network-level spiking activity while producing classification accuracy that are comparable to standard approaches on the same dataset.

An Inkjet-Printed PEDOT:PSS-Based Stretchable Conductor for Wearable Health Monitoring Device Applications.

A stretchable conductor is one of the key components in soft electronics that allows the seamless integration of electronic devices and sensors on elastic substrates. Its unique advantages of mechanical flexibility and stretchability have enabled a variety of wearable bioelectronic devices that can conformably adapt to curved skin surfaces for long-term health monitoring applications. Here, we report a poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS)-based stretchable polymer blend that can be patterned using an inkjet printing process while exhibiting low sheet resistance and accommodating large mechanical deformations. We have systematically studied the effect of various types of polar solvent additives that can help induce phase separation of PEDOT and PSS grains and change the conformation of a PEDOT chain, thereby improving the electrical property of the film by facilitating charge hopping along the percolating PEDOT network. The optimal ink formulation is achieved by adding 5 wt % ethylene glycol into a pristine PEDOT:PSS aqueous solution, which results in a sheet resistance of as low as 58 Ω/□. Elasticity can also be achieved by blending the above solution with the soft polymer poly(ethylene oxide) (PEO). Thin films of PEDOT:PSS/PEO polymer blends patterned by inkjet printing exhibits a low sheet resistance of 84 Ω/□ and can resist up to 50% tensile strain with minimal changes in electrical performance. With its good conductivity and elasticity, we have further demonstrated the use of the polymer blend as stretchable interconnects and stretchable dry electrodes on a thin polydimethylsiloxane (PDMS) substrate for photoplethysmography (PPG) and electrocardiography (ECG) recording applications. This work shows the potential of using a printed stretchable conducting polymer in low-cost wearable sensor patches for smart health applications.

Sub-Nanowatt Ultrasonic Bio-Telemetry Using B-Scan Imaging.

GOAL: The objective of this paper is to investigate if the use of a B-scan ultrasound imaging system can reduce the energy requirements, and hence the power-dissipation requirements to support wireless bio-telemetry at an implantable device. METHODS: B-scan imaging data were acquired using a commercial 256-element linear ultrasound transducer array which was driven by a commercial echoscope. As a transmission medium, we used a water-bath and the operation of the implantable device was emulated using a commercial-off-the-shelf micro-controller board. The telemetry parameters (e.g. transmission rate and transmission power) were wirelessly controlled using a two-way radio-frequency transceiver. B-scan imaging data were post-processed using a maximum-threshold decoder and the quality of the ultrasonic telemetry link was quantified in terms of its bit-error-rate (BER). RESULTS: Measured results show that a reliable B-scan communication link with an implantable device can be achieved at transmission power levels of 100 pW and for implantation depths greater than 10 cm. CONCLUSIONS: In this paper we demonstrated that a combination of B-scan imaging and a simple decoding algorithm can significantly reduce the energy-budget requirements for reliable ultrasonic telemetry.

Resonant Machine Learning Based on Complex Growth Transform Dynamical Systems.

Traditional energy-based learning models associate a single energy metric to each configuration of variables involved in the underlying optimization process. Such models associate the lowest energy state with the optimal configuration of variables under consideration and are thus inherently dissipative. In this article, we propose an energy-efficient learning framework that exploits structural and functional similarities between a machine-learning network and a general electrical network satisfying Tellegen's theorem. In contrast to the standard energy-based models, the proposed formulation associates two energy components, namely, active and reactive energy with the network. The formulation ensures that the network's active power is dissipated only during the process of learning, whereas the reactive power is maintained to be zero at all times. As a result, in steady state, the learned parameters are stored and self-sustained by electrical resonance determined by the network's nodal inductances and capacitances. Based on this approach, this article introduces three novel concepts: 1) a learning framework where the network's active-power dissipation is used as a regularization for a learning objective function that is subjected to zero total reactive-power constraint; 2) a dynamical system based on complex-domain, continuous-time growth transforms that optimizes the learning objective function and drives the network toward electrical resonance under steady-state operation; and 3) an annealing procedure that controls the tradeoff between active-power dissipation and the speed of convergence. As a representative example, we show how the proposed framework can be used for designing resonant support vector machines (SVMs), where the support vectors correspond to an LC network with self-sustained oscillations. We also show that this resonant network dissipates less active power compared with its non-resonant counterpart.

A self-powered analog sensor-data-logging device based on Fowler-Nordheim dynamical systems.

Continuous, battery-free operation of sensor nodes requires ultra-low-power sensing and data-logging techniques. Here we report that by directly coupling a sensor/transducer signal into globally asymptotically stable monotonic dynamical systems based on Fowler-Nordheim quantum tunneling, one can achieve self-powered sensing at an energy budget that is currently unachievable using conventional energy harvesting methods. The proposed device uses a differential architecture to compensate for environmental variations and the device can retain sensed information for durations ranging from hours to days. With a theoretical operating energy budget less than 10 attojoules, we demonstrate that when integrated with a miniature piezoelectric transducer the proposed sensor-data-logger can measure cumulative "action" due to ambient mechanical acceleration without any additional external power.

A Spiking Neuron and Population Model Based on the Growth Transform Dynamical System.

In neuromorphic engineering, neural populations are generally modeled in a bottom-up manner, where individual neuron models are connected through synapses to form large-scale spiking networks. Alternatively, a top-down approach treats the process of spike generation and neural representation of excitation in the context of minimizing some measure of network energy. However, these approaches usually define the energy functional in terms of some statistical measure of spiking activity (ex. firing rates), which does not allow independent control and optimization of neurodynamical parameters. In this paper, we introduce a new spiking neuron and population model where the dynamical and spiking responses of neurons can be derived directly from a network objective or energy functional of continuous-valued neural variables like the membrane potential. The key advantage of the model is that it allows for independent control over three neuro-dynamical properties: (a) control over the steady-state population dynamics that encodes the minimum of an exact network energy functional; (b) control over the shape of the action potentials generated by individual neurons in the network without affecting the network minimum; and (c) control over spiking statistics and transient population dynamics without affecting the network minimum or the shape of action potentials. At the core of the proposed model are different variants of Growth Transform dynamical systems that produce stable and interpretable population dynamics, irrespective of the network size and the type of neuronal connectivity (inhibitory or excitatory). In this paper, we present several examples where the proposed model has been configured to produce different types of single-neuron dynamics as well as population dynamics. In one such example, the network is shown to adapt such that it encodes the steady-state solution using a reduced number of spikes upon convergence to the optimal solution. In this paper, we use this network to construct a spiking associative memory that uses fewer spikes compared to conventional architectures, while maintaining high recall accuracy at high memory loads.

Refreshable Nanobiosensor Based on Organosilica Encapsulation of Biorecognition Elements.

Implantable and wearable biosensors that enable monitoring of biophysical and biochemical parameters over long durations are highly attractive for early and presymptomatic diagnosis of pathological conditions and timely clinical intervention. Poor stability of antibodies used as biorecognition elements and the lack of effective methods to refresh the biosensors upon demand without severely compromising the functionality of the biosensor remain significant challenges in realizing protein biosensors for long-term monitoring. Here, we introduce a novel method involving organosilica encapsulation of antibodies for preserving their biorecognition capability under harsh conditions, typically encountered during the sensor refreshing process, and elevated temperature. Specifically, a simple aqueous rinsing step using sodium dodecyl sulfate (SDS) solution refreshes the biosensor by dissociating the antibody-antigen interactions. Encapsulation of the antibodies with an organosilica layer is shown to preserve the biorecognition capability of otherwise unstable antibodies during the SDS treatment, thus ultimately facilitating the refreshability of the biosensor over multiple cycles. Harnessing this method, we demonstrate the refreshability of plasmonic biosensors for anti-IgG (model bioanalyte) and neutrophil gelatinase-associated lipocalin (NGAL) (a biomarker for acute and chronic kidney injury). The novel encapsulation approach demonstrated can be easily extended to other transduction platforms to realize refreshable biosensors for monitoring of protein biomarkers over long durations.

Exploiting Self-Capacitances for Wireless Power Transfer.

Conventional approaches for wireless power transfer rely on the mutual coupling (near-field or far-field) between the transmitter and receiver transducers. As a result, the power-transfer efficiency of these approaches scales non-linearly with the cross-sectional area of the transducers and with the relative distance and respective alignment between the transducers. In this paper, we show that when the operational power-budget requirements are in the order of microwatts, a self-capacitance (SC)-based power delivery has significant advantages in terms of the power transfer-efficiency, receiver form-factor, and system scalability when compared to other modes of wireless power transfer (WPT) methods. We present a simple and a tractable equivalent circuit model that can be used to study the effect of different parameters on the SC-based WPT. In this paper, we have experimentally verified the validity of the circuit using a cadaver mouse model. We also demonstrate the feasibility of a hybrid telemetry system where the microwatts of power, which can be harvested from SC-based WPT approach, is used for back-scattering a radio-frequency (RF) signal and is used for remote sensing of in vivo physiological parameters such as temperature. The functionality of the hybrid system has also been verified using a cadaver mouse model housed in a cage that was retrofitted with 915 MHz RF back-scattering antennas. We believe that the proposed remote power-delivery and hybrid telemetry approach would be useful in remote activation of wearable devices and in the design of energy-efficient animal cages used for long-term monitoring applications.

Differential Fowler-Nordheim Tunneling Dynamical System for Attojoule Sensing and Recording.

Dynamical systems that evolve unidirectionally with respect to time provide a natural mechanism for implementing a time-domain, near-zero-threshold energy rectifier. In this paper we implement such a dynamical system using a pair of differential, leaky floating-gates and demonstrate that the circuit can sense and record signals of interest while compensating for environmental variations. A Fowler-Nordheim (FN) tunneling current has been used to implement the leakage process, which we experimentally show can be modulated by signals at energy levels below femtojoules. At this level of energy, the proposed FN-system could be self-powered using different types of biopotential energy sources like intra-cellular potentials, a feature that was not possible with previously reported recorders. Furthermore, the degree of modulation is shown to be a function of the input intensity as well as time-of-occurrence, which opens up the possibility of using reconstruction techniques to reconstruct the input signal from measurement of multiple sensing devices. Using devices fabricated in a 0.5 µm standard CMOS process, we demonstrate recording of 6 mV events with retention capability lasting over 30 minutes.

Feasibility of Self-Powering and Energy Harvesting Using Cardiac Valvular Perturbations.

In this paper, we investigate the feasibility of harvesting energy from cardiac valvular perturbations to self-power a wireless sonomicrometry sensor. Compared to the previous studies involving piezoelectric patches or encasings attached to the cardiac or aortic surface, the proposed study explores the use of piezoelectric sutures that can be implanted in proximity to the valvular regions, where non-linear valvular perturbations could be exploited for self-powering. Using an ovine animal model, the magnitude of valvular perturbations are first measured using an array of sonomicrometry crystals implanted around the tricuspid valve. These measurements were then used to estimate the levels of electrical energy that could be harvested using a simplified piezoelectric suture model. These results were revalidated across seven different animals, before and after valvular regurgitation was induced. Our study shows that power harvested from different annular planes of the tricuspid valve (before and after regurgitation) could range from nano-watts to milli-watts, with the maximum power harvested from the leaflet plane. We believe that these results could be useful for determining optimal surgical placement of wireless and self-powered sonomicrometry sensor, which in turn could be used for investigating the pathophysiology of ischemic regurgitation.

Decentralized Global Optimization Based on a Growth Transform Dynamical System Model.

Conservation principles, such as conservation of charge, energy, or mass, provide a natural way to couple and constrain spatially separated variables. In this paper, we propose a dynamical system model that exploits these constraints for solving nonconvex and discrete global optimization problems. Unlike the traditional simulated annealing or quantum annealing-based global optimization techniques, the proposed method optimizes a target objective function by continuously evolving a driver functional over a conservation manifold, using a generalized variant of growth transformations. As a result, the driver functional asymptotically converges toward a Dirac-delta function that is centered at the global optimum of the target objective function. In this paper, we provide an outline of the proof of convergence for the dynamical system model and investigate different properties of the model using a benchmark nonlinear optimization problem. Also, we demonstrate how a discrete variant of the proposed dynamical system can be used for implementing decentralized optimization algorithms, where an ensemble of spatially separated entities (for example, biological cells or simple computational units) can collectively implement specific functions, such as winner-take-all and ranking, by exchanging signals only with its immediate substrate or environment. The proposed dynamical system model could potentially be used to implement continuous-time optimizers, annealers, and neural networks.