Search

There is increasing interest in the application large language models (LLMs) to the medical field, in part because of their impressive performance on medical exam questions. While promising, exam questions do not reflect the complexity of real patient-doctor interactions. In reality, physicians' decisions are shaped by many complex factors, such as patient compliance, personal experience, ethical beliefs, and cognitive bias. Taking a step toward understanding this, our hypothesis posits that when LLMs are confronted with clinical questions containing cognitive biases, they will yield significantly less accurate responses compared to the same questions presented without such biases. In this study, we developed BiasMedQA, a benchmark for evaluating cognitive biases in LLMs applied to medical tasks. Using BiasMedQA we evaluated six LLMs, namely GPT-4, Mixtral-8x70B, GPT-3.5, PaLM-2, Llama 2 70B-chat, and the medically specialized PMC Llama 13B. We tested these models on 1,273 questions from the US Medical Licensing Exam (USMLE) Steps 1, 2, and 3, modified to replicate common clinically-relevant cognitive biases. Our analysis revealed varying effects for biases on these LLMs, with GPT-4 standing out for its resilience to bias, in contrast to Llama 2 70B-chat and PMC Llama 13B, which were disproportionately affected by cognitive bias. Our findings highlight the critical need for bias mitigation in the development of medical LLMs, pointing towards safer and more reliable applications in healthcare.

Matrix multiplication (MatMul) typically dominates the overall computational cost of large language models (LLMs). This cost only grows as LLMs scale to larger embedding dimensions and context lengths. In this work, we show that MatMul operations can be completely eliminated from LLMs while maintaining strong performance at billion-parameter scales. Our experiments show that our proposed MatMul-free models achieve performance on-par with state-of-the-art Transformers that require far more memory during inference at a scale up to at least 2.7B parameters. We investigate the scaling laws and find that the performance gap between our MatMul-free models and full precision Transformers narrows as the model size increases. We also provide a GPU-efficient implementation of this model which reduces memory usage by up to 61% over an unoptimized baseline during training. By utilizing an optimized kernel during inference, our model's memory consumption can be reduced by more than 10x compared to unoptimized models. To properly quantify the efficiency of our architecture, we build a custom hardware solution on an FPGA which exploits lightweight operations beyond what GPUs are capable of. We processed billion-parameter scale models at 13W beyond human readable throughput, moving LLMs closer to brain-like efficiency. This work not only shows how far LLMs can be stripped back while still performing effectively, but also points at the types of operations future accelerators should be optimized for in processing the next generation of lightweight LLMs. Our code implementation is available at \url{https://github.com/ridgerchu/matmulfreellm}.

Autonomous driving demands an integrated approach that encompasses perception, prediction, and planning, all while operating under strict energy constraints to enhance scalability and environmental sustainability. We present Spiking Autonomous Driving (SAD), the first unified Spiking Neural Network (SNN) to address the energy challenges faced by autonomous driving systems through its event-driven and energy-efficient nature. SAD is trained end-to-end and consists of three main modules: perception, which processes inputs from multi-view cameras to construct a spatiotemporal bird's eye view; prediction, which utilizes a novel dual-pathway with spiking neurons to forecast future states; and planning, which generates safe trajectories considering predicted occupancy, traffic rules, and ride comfort. Evaluated on the nuScenes dataset, SAD achieves competitive performance in perception, prediction, and planning tasks, while drawing upon the energy efficiency of SNNs. This work highlights the potential of neuromorphic computing to be applied to energy-efficient autonomous driving, a critical step toward sustainable and safety-critical automotive technology. Our code is available at \url{https://github.com/ridgerchu/SAD}.

Recent advancements in neuroscience research have propelled the development of Spiking Neural Networks (SNNs), which not only have the potential to further advance neuroscience research but also serve as an energy-efficient alternative to Artificial Neural Networks (ANNs) due to their spike-driven characteristics. However, previous studies often neglected the multiscale information and its spatiotemporal correlation between event data, leading SNN models to approximate each frame of input events as static images. We hypothesize that this oversimplification significantly contributes to the performance gap between SNNs and traditional ANNs. To address this issue, we have designed a Spiking Multiscale Attention (SMA) module that captures multiscale spatiotemporal interaction information. Furthermore, we developed a regularization method named Attention ZoneOut (AZO), which utilizes spatiotemporal attention weights to reduce the model's generalization error through pseudo-ensemble training. Our approach has achieved state-of-the-art results on mainstream neural morphology datasets. Additionally, we have reached a performance of 77.1% on the Imagenet-1K dataset using a 104-layer ResNet architecture enhanced with SMA and AZO. This achievement confirms the state-of-the-art performance of SNNs with non-transformer architectures and underscores the effectiveness of our method in bridging the performance gap between SNN models and traditional ANN models.

Weight quantization is used to deploy high-performance deep learning models on resource-limited hardware, enabling the use of low-precision integers for storage and computation. Spiking neural networks (SNNs) share the goal of enhancing efficiency, but adopt an 'event-driven' approach to reduce the power consumption of neural network inference. While extensive research has focused on weight quantization, quantization-aware training (QAT), and their application to SNNs, the precision reduction of state variables during training has been largely overlooked, potentially diminishing inference performance. This paper introduces two QAT schemes for stateful neurons: (i) a uniform quantization strategy, an established method for weight quantization, and (ii) threshold-centered quantization, which allocates exponentially more quantization levels near the firing threshold. Our results show that increasing the density of quantization levels around the firing threshold improves accuracy across several benchmark datasets. We provide an ablation analysis of the effects of weight and state quantization, both individually and combined, and how they impact models. Our comprehensive empirical evaluation includes full precision, 8-bit, 4-bit, and 2-bit quantized SNNs, using QAT, stateful QAT (SQUAT), and post-training quantization methods. The findings indicate that the combination of QAT and SQUAT enhance performance the most, but given the choice of one or the other, QAT improves performance by the larger degree. These trends are consistent all datasets. Our methods have been made available in our Python library snnTorch: https://github.com/jeshraghian/snntorch., Comment: 10 pages, 4 figures, accepted at NICE 2024

The surge of remote and hybrid work in the post-pandemic era has reinforced the blurred boundaries between work and nonwork responsibilities. Thus, how people manage the boundaries between work and nonwork domains has become more complicated. This study advances the work of previous studies on constant connectivity by focusing on how employees’ perception of constant connectivity might actualize their boundary management behaviors. By adopting affordances for practice perspective, our study focused on contextual factors, including IT/internet policies, informal social norms, and work flexibility, to investigate how these factors could influence employees’ perception of constant connectivity. This paper reports a two-phase study. In the first phase, we used sentiment analysis to rank 38 internet use policies of Australian universities, grading their strictness toward ICT/internet use. Next, building on the first phase, we interviewed 28 academics. We identified three perceptions of constant connectivity related to participants’ practices, including constant connectivity as a resource for practice, a challenge for practice, and duality for practice. We also found five distinct boundary management behaviors connected to three different perceptions.

The field of neuromorphic computing holds great promise in terms of advancing computing efficiency and capabilities by following brain-inspired principles. However, the rich diversity of techniques employed in neuromorphic research has resulted in a lack of clear standards for benchmarking, hindering effective evaluation of the advantages and strengths of neuromorphic methods compared to traditional deep-learning-based methods. This paper presents a collaborative effort, bringing together members from academia and the industry, to define benchmarks for neuromorphic computing: NeuroBench. The goals of NeuroBench are to be a collaborative, fair, and representative benchmark suite developed by the community, for the community. In this paper, we discuss the challenges associated with benchmarking neuromorphic solutions, and outline the key features of NeuroBench. We believe that NeuroBench will be a significant step towards defining standards that can unify the goals of neuromorphic computing and drive its technological progress. Please visit neurobench.ai for the latest updates on the benchmark tasks and metrics.

As the size of large language models continue to scale, so does the computational resources required to run it. Spiking Neural Networks (SNNs) have emerged as an energy-efficient approach to deep learning that leverage sparse and event-driven activations to reduce the computational overhead associated with model inference. While they have become competitive with non-spiking models on many computer vision tasks, SNNs have also proven to be more challenging to train. As a result, their performance lags behind modern deep learning, and we are yet to see the effectiveness of SNNs in language generation. In this paper, inspired by the Receptance Weighted Key Value (RWKV) language model, we successfully implement `SpikeGPT', a generative language model with binary, event-driven spiking activation units. We train the proposed model on two model variants: 45M and 216M parameters. To the best of our knowledge, SpikeGPT is the largest backpropagation-trained SNN model to date, rendering it suitable for both the generation and comprehension of natural language. We achieve this by modifying the transformer block to replace multi-head self attention to reduce quadratic computational complexity O(N^2) to linear complexity O(N) with increasing sequence length. Input tokens are instead streamed in sequentially to our attention mechanism (as with typical SNNs). Our preliminary experiments show that SpikeGPT remains competitive with non-spiking models on tested benchmarks, while maintaining 20x fewer operations when processed on neuromorphic hardware that can leverage sparse, event-driven activations.

This paper presents a spiking neural network (SNN) accelerator made using fully open-source EDA tools, process design kit (PDK), and memory macros synthesized using OpenRAM. The chip is taped out in the 130 nm SkyWater process and integrates over 1 million synaptic weights, and offers a reprogrammable architecture. It operates at a clock speed of 40 MHz, a supply of 1.8 V, uses a PicoRV32 core for control, and occupies an area of 33.3 mm^2. The throughput of the accelerator is 48,262 images per second with a wallclock time of 20.72 us, at 56.8 GOPS/W. The spiking neurons use hysteresis to provide an adaptive threshold (i.e., a Schmitt trigger) which can reduce state instability. This results in high performing SNNs across a range of benchmarks that remain competitive with state-of-the-art, full precision SNNs. The design is open sourced and available online: https://github.com/sfmth/OpenSpike, Comment: The design is open sourced and available online: https://github.com/sfmth/OpenSpike

Artificial neural networks (ANNs) have emerged as an essential tool in machine learning, achieving remarkable success across diverse domains, including image and speech generation, game playing, and robotics. However, there exist fundamental differences between ANNs' operating mechanisms and those of the biological brain, particularly concerning learning processes. This paper presents a comprehensive review of current brain-inspired learning representations in artificial neural networks. We investigate the integration of more biologically plausible mechanisms, such as synaptic plasticity, to enhance these networks' capabilities. Moreover, we delve into the potential advantages and challenges accompanying this approach. Ultimately, we pinpoint promising avenues for future research in this rapidly advancing field, which could bring us closer to understanding the essence of intelligence.

Analog compute-in-memory (CIM) systems are promising for deep neural network (DNN) inference acceleration due to their energy efficiency and high throughput. However, as the use of DNNs expands, protecting user input privacy has become increasingly important. In this paper, we identify a potential security vulnerability wherein an adversary can reconstruct the user's private input data from a power side-channel attack, under proper data acquisition and pre-processing, even without knowledge of the DNN model. We further demonstrate a machine learning-based attack approach using a generative adversarial network (GAN) to enhance the data reconstruction. Our results show that the attack methodology is effective in reconstructing user inputs from analog CIM accelerator power leakage, even at large noise levels and after countermeasures are applied. Specifically, we demonstrate the efficacy of our approach on an example of U-Net inference chip for brain tumor detection, and show the original magnetic resonance imaging (MRI) medical images can be successfully reconstructed even at a noise-level of 20% standard deviation of the maximum power signal value. Our study highlights a potential security vulnerability in analog CIM accelerators and raises awareness of using GAN to breach user privacy in such systems.

Neuromorphic computing shows promise for advancing computing efficiency and capabilities of AI applications using brain-inspired principles. However, the neuromorphic research field currently lacks standardized benchmarks, making it difficult to accurately measure technological advancements, compare performance with conventional methods, and identify promising future research directions. Prior neuromorphic computing benchmark efforts have not seen widespread adoption due to a lack of inclusive, actionable, and iterative benchmark design and guidelines. To address these shortcomings, we present NeuroBench: a benchmark framework for neuromorphic computing algorithms and systems. NeuroBench is a collaboratively-designed effort from an open community of nearly 100 co-authors across over 50 institutions in industry and academia, aiming to provide a representative structure for standardizing the evaluation of neuromorphic approaches. The NeuroBench framework introduces a common set of tools and systematic methodology for inclusive benchmark measurement, delivering an objective reference framework for quantifying neuromorphic approaches in both hardware-independent (algorithm track) and hardware-dependent (system track) settings. In this article, we present initial performance baselines across various model architectures on the algorithm track and outline the system track benchmark tasks and guidelines. NeuroBench is intended to continually expand its benchmarks and features to foster and track the progress made by the research community., Comment: Updated from whitepaper to full perspective article preprint

Imaging flow cytometry (FC) is a powerful analytic tool that combines the principles of conventional FC with rich spatial information, allowing more profound insight into single-cell analysis. However, offering such high-resolution, full-frame feedback can restrain processing speed and has become a significant trade-off during development. In addition, the dynamic range (DR) offered by conventional photosensors can only capture limited fluorescence signals, which compromises the detection of high-velocity fluorescent objects. Neuromorphic photo-sensing focuses on the events of interest via individual-firing pixels to reduce data redundancy and latency. With its inherent high DR, this architecture has the potential to drastically elevate the performance in throughput and sensitivity to fluorescent targets. Herein, we presented an early demonstration of neuromorphic cytometry, demonstrating the feasibility of adopting an event-based resolution in describing spatiotemporal feedback on microscale objects and for the first time, including cytometric-like functions in object counting and size estimation to measure 8 µm, 15 µm microparticles and human monocytic cell line (THP-1). Our work has achieved highly consistent outputs with a widely adopted flow cytometer (CytoFLEX) in detecting microparticles. Moreover, the capacity of an event-based photosensor in registering fluorescent signals was evaluated by recording 6 µm Fluorescein isothiocyanate-marked particles in different lighting conditions, revealing superior performance compared to a standard photosensor. Although the current platform cannot deliver multiparametric measurements on cells, future endeavours will include further functionalities and increase the measurement parameters (granularity, cell condition, fluorescence analysis) to enrich cell interpretation.

Spiking neural networks and neuromorphic hardware platforms that emulate neural dynamics are slowly gaining momentum and entering main-stream usage. Despite a well-established mathematical foundation for neural dynamics, the implementation details vary greatly across different platforms. Correspondingly, there are a plethora of software and hardware implementations with their own unique technology stacks. Consequently, neuromorphic systems typically diverge from the expected computational model, which challenges the reproducibility and reliability across platforms. Additionally, most neuromorphic hardware is limited by its access via a single software frameworks with a limited set of training procedures. Here, we establish a common reference-frame for computations in neuromorphic systems, dubbed the Neuromorphic Intermediate Representation (NIR). NIR defines a set of computational primitives as idealized continuous-time hybrid systems that can be composed into graphs and mapped to and from various neuromorphic technology stacks. By abstracting away assumptions around discretization and hardware constraints, NIR faithfully captures the fundamental computation, while simultaneously exposing the exact differences between the evaluated implementation and the idealized mathematical formalism. We reproduce three NIR graphs across 7 neuromorphic simulators and 4 hardware platforms, demonstrating support for an unprecedented number of neuromorphic systems. With NIR, we decouple the evolution of neuromorphic hardware and software, ultimately increasing the interoperability between platforms and improving accessibility to neuromorphic technologies. We believe that NIR is an important step towards the continued study of brain-inspired hardware and bottom-up approaches aimed at an improved understanding of the computational underpinnings of nervous systems., Comment: NIR is available at https://github.com/neuromorphs/NIR

Recent advances in robot-assisted surgery have resulted in progressively more precise, efficient, and minimally invasive procedures, sparking a new era of robotic surgical intervention. This enables doctors, in collaborative interaction with robots, to perform traditional or minimally invasive surgeries with improved outcomes through smaller incisions. Recent efforts are working toward making robotic surgery more autonomous which has the potential to reduce variability of surgical outcomes and reduce complication rates. Deep reinforcement learning methodologies offer scalable solutions for surgical automation, but their effectiveness relies on extensive data acquisition due to the absence of prior knowledge in successfully accomplishing tasks. Due to the intensive nature of simulated data collection, previous works have focused on making existing algorithms more efficient. In this work, we focus on making the simulator more efficient, making training data much more accessible than previously possible. We introduce Surgical Gym, an open-source high performance platform for surgical robot learning where both the physics simulation and reinforcement learning occur directly on the GPU. We demonstrate between 100-5000x faster training times compared with previous surgical learning platforms. The code is available at: https://github.com/SamuelSchmidgall/SurgicalGym.

Diagnostic investigation has an important role in risk stratification and clinical decision making of patients with suspected and documented Coronary Artery Disease (CAD). However, the majority of existing tools are primarily focused on the selection of gatekeeper tests, whereas only a handful of systems contain information regarding the downstream testing or treatment. We propose a multi-task deep learning model to support risk stratification and down-stream test selection for patients undergoing Coronary Computed Tomography Angiography (CCTA). The analysis included 14,021 patients who underwent CCTA between 2006 and 2017. Our novel multitask deep learning framework extends the state-of-the art Perceiver model to deal with real-world CCTA report data. Our model achieved an Area Under the receiver operating characteristic Curve (AUC) of 0.76 in CAD risk stratification, and 0.72 AUC in predicting downstream tests. Our proposed deep learning model can accurately estimate the likelihood of CAD and provide recommended downstream tests based on prior CCTA data. In clinical practice, the utilization of such an approach could bring a paradigm shift in risk stratification and downstream management. Despite significant progress using deep learning models for tabular data, they do not outperform gradient boosting decision trees, and further research is required in this area. However, neural networks appear to benefit more readily from multi-task learning than tree-based models. This could offset the shortcomings of using single task learning approach when working with tabular data.

Neuromodulation techniques have emerged as promising approaches for treating a wide range of neurological disorders, precisely delivering electrical stimulation to modulate abnormal neuronal activity. While leveraging the unique capabilities of artificial intelligence (AI) holds immense potential for responsive neurostimulation, it appears as an extremely challenging proposition where real-time (low-latency) processing, low power consumption, and heat constraints are limiting factors. The use of sophisticated AI-driven models for personalized neurostimulation depends on back-telemetry of data to external systems (e.g. cloud-based medical mesosystems and ecosystems). While this can be a solution, integrating continuous learning within implantable neuromodulation devices for several applications, such as seizure prediction in epilepsy, is an open question. We believe neuromorphic architectures hold an outstanding potential to open new avenues for sophisticated on-chip analysis of neural signals and AI-driven personalized treatments. With more than three orders of magnitude reduction in the total data required for data processing and feature extraction, the high power- and memory-efficiency of neuromorphic computing to hardware-firmware co-design can be considered as the solution-in-the-making to resource-constraint implantable neuromodulation systems. This perspective introduces the concept of Neuromorphic Neuromodulation, a new breed of closed-loop responsive feedback system. It highlights its potential to revolutionize implantable brain-machine microsystems for patient-specific treatment

As deep learning models scale, they become increasingly competitive from domains spanning from computer vision to natural language processing; however, this happens at the expense of efficiency since they require increasingly more memory and computing power. The power efficiency of the biological brain outperforms any large-scale deep learning ( DL ) model; thus, neuromorphic computing tries to mimic the brain operations, such as spike-based information processing, to improve the efficiency of DL models. Despite the benefits of the brain, such as efficient information transmission, dense neuronal interconnects, and the co-location of computation and memory, the available biological substrate has severely constrained the evolution of biological brains. Electronic hardware does not have the same constraints; therefore, while modeling spiking neural networks ( SNNs) might uncover one piece of the puzzle, the design of efficient hardware backends for SNN s needs further investigation, potentially taking inspiration from the available work done on the artificial neural networks ( ANNs) side. As such, when is it wise to look at the brain while designing new hardware, and when should it be ignored? To answer this question, we quantitatively compare the digital hardware acceleration techniques and platforms of ANNs and SNN s. As a result, we provide the following insights: (i) ANNs currently process static data more efficiently, (ii) applications targeting data produced by neuromorphic sensors, such as event-based cameras and silicon cochleas, need more investigation since the behavior of these sensors might naturally fit the SNN paradigm, and (iii) hybrid approaches combining SNN s and ANNs might lead to the best solutions and should be investigated further at the hardware level, accounting for both efficiency and loss optimization., Comment: Fixed error in bio

Memristive reservoirs draw inspiration from a novel class of neuromorphic hardware known as nanowire networks. These systems display emergent brain-like dynamics, with optimal performance demonstrated at dynamical phase transitions. In these networks, a limited number of electrodes are available to modulate system dynamics, in contrast to the global controllability offered by neuromorphic hardware through random access memories. We demonstrate that the learn-to-learn framework can effectively address this challenge in the context of optimization. Using the framework, we successfully identify the optimal hyperparameters for the reservoir. This finding aligns with previous research, which suggests that the optimal performance of a memristive reservoir occurs at the `edge of formation' of a conductive pathway. Furthermore, our results show that these systems can mimic membrane potential behavior observed in spiking neurons, and may serve as an interface between spike-based and continuous processes., Comment: 7 pages, 6 figures, ICONS 2023, accepted

The obesity paradox is characterized by a U- or J-shaped curve, suggesting that overweight individuals have a lower risk of mortality compared to their normal weight counterparts. It is largely unknown whether muscle density plays a role in the relationship between BMI and mortality. We use data from the Multi-Ethnic Study of Atherosclerosis, an ethnically diverse longitudinal study of males and females, to examine whether muscle density or sex modify the BMI-mortality relationship. Results indicate that obese males with high muscle density have a lower risk of mortality (HR: 0.59, 95% CI: (0.23, 1.50)) compared to their normal weight counterparts, while obese females with high muscle density have a higher risk of mortality (HR: 1.37, 95% CI: 0.63, 3.00)) compared to their normal weight counterparts. Our results show that muscle density is a mediator, but not an effect modifier, in the association between BMI and mortality. Sex is an effect modifier in this association, with attenuated risk of mortality among obese men (HR: 0.50, 95% CI: (0.36, 0.69)), but not obese women (HR: 0.95, 95% CI: (0.65, 1.39)). Overall, our data are consistent with the survival advantage seen in obese populations.

The obesity paradox is characterized by a U- or J-shaped curve, suggesting that overweight individuals have a lower risk of mortality compared to their normal weight counterparts. It is largely unknown whether muscle density plays a role in the relationship between BMI and mortality. We use data from the Multi-Ethnic Study of Atherosclerosis, an ethnically diverse longitudinal study of males and females, to examine whether muscle density or sex modify the BMI-mortality relationship. Results indicate that obese males with high muscle density have a lower risk of mortality (HR: 0.59, 95% CI: (0.23, 1.50)) compared to their normal weight counterparts, while obese females with high muscle density have a higher risk of mortality (HR: 1.37, 95% CI: 0.63, 3.00)) compared to their normal weight counterparts. Our results show that muscle density is a mediator, but not an effect modifier, in the association between BMI and mortality. Sex is an effect modifier in this association, with attenuated risk of mortality among obese men (HR: 0.50, 95% CI: (0.36, 0.69)), but not obese women (HR: 0.95, 95% CI: (0.65, 1.39)). Overall, our data are consistent with the survival advantage seen in obese populations.

HIV prevention and control methods are implemented on different scales to reduce the spread of the virus amongst populations. However, despite such efforts, HIV continues to persist in populations with a global incidence rate of 1.8 million in 2017 alone. The introduction of new infections into susceptible regional populations promotes the spread of HIV, indicating a crucial need to study the impact of migration and mobility on regional and global efforts to prevent HIV transmission. Here we reviewed studies that assess the impact of human mobility on HIV transmission and spread. We found an important role for both travel and migration in driving the spread of HIV across regional and national borders. Combined, our results indicate that even in the presence of control and preventive efforts, if migration and travel are occurring, public health efforts will need to remain persistent to ensure that new infections do not grow into outbreaks.

This dissertation is dedicated to using Memristive Spiking Neural Networks (MSNNs) for deep learning tasks such as image classification, visual associative memory tasks such as pattern recognition, and auditory cortex processing tasks such as sound localization (SL). The image classification model consists entirely of memristive neurons and memristive synapses utilizing deep learning plausible supervised learning rules. The pattern recognition fully MSNNs consists of memristive neurons and memristive synapses harnessing biologically plausible unsupervised learning rules. SL MSNNs emulate biological brain functionality with volatile memristive synapses.By developing a minimal circuit element memristive neuron -- Memristive Integrate-and-Fire (MIF) neuron -- with commercially accessible memristors, we are able to demonstrate large-scale fully MSNNs and apply the supervised Backpropagation Through Time (BPTT) algorithm to train networks, achieving state-of-the-art accuracy for several datasets. In addition, using a memristive unsupervised learning rule based on a continuously evolving alpha function membrane potential, we are also able to train large-scale fully MSNNs with Spiking-Time-Dependent Plasticity (STDP) and achieve high accuracies. Moreover, a volatile memristor is used for mimicking Short-Term Depression (STD) synapses such that we can simulate SL networks and pinpoint the direction of the sound coming from.The major contributions reported in this proposal include: • Development of the MIF neuron SPICE-level model; • Validation of the SPICE-level MIF neuron model to a Python-based simulation of large-scale MSNNs; • Simulation of a small-scale neuron network consisting of a presynaptic, a memristive synapse, and a postsynaptic MIF neuron to generate the STDP learning window; • Abstraction of memristive STDP with alpha functions; • Simulation of a large-scale, fully MSNN consisting of MIF neurons and memristive synapses for unsupervised learning in Python; • D

This paper proposes an artificial intelligence system that continuously improves over time at event prediction using initially unlabelled data by using self-supervised learning. Time-series data are inherently autocorrelated. By using a detection model to generate weak labels on the fly, which are concurrently used as targets to train a prediction model on a time-shifted input data stream, this autocorrelation can effectively be harnessed to reduce the burden of manual labelling. This is critical in medical patient monitoring, as it enables the development of personalized forecasting models without demanding the annotation of long sequences of physiological signal recordings. We perform a feasibility study on seizure prediction, which is identified as an ideal test case, as pre-ictal brainwaves are patient-specific, and tailoring models to individual patients is known to improve forecasting performance significantly. Our self-supervised approach is used to train individualized forecasting models for 10 patients, showing an average relative improvement in sensitivity by 14.30% and a reduction in false alarms by 19.61% in early seizure forecasting. This proof-of-concept on the feasibility of using a continuous stream of time-series neurophysiological data paves the way towards a low-power neuromorphic neuromodulation system.

Spiking neural networks (SNNs) have achieved orders of magnitude improvement in terms of energy consumption and latency when performing inference with deep learning workloads. Error backpropagation is presently regarded as the most effective method for training SNNs, but in a twist of irony, when training on modern graphics processing units (GPUs) this becomes more expensive than non-spiking networks. The emergence of Graphcore's Intelligence Processing Units (IPUs) balances the parallelized nature of deep learning workloads with the sequential, reusable, and sparsified nature of operations prevalent when training SNNs. IPUs adopt multi-instruction multi-data (MIMD) parallelism by running individual processing threads on smaller data blocks, which is a natural fit for the sequential, non-vectorized steps required to solve spiking neuron dynamical state equations. We present an IPU-optimized release of our custom SNN Python package, snnTorch, which exploits fine-grained parallelism by utilizing low-level, pre-compiled custom operations to accelerate irregular and sparse data access patterns that are characteristic of training SNN workloads. We provide a rigorous performance assessment across a suite of commonly used spiking neuron models, and propose methods to further reduce training run-time via half-precision training. By amortizing the cost of sequential processing into vectorizable population codes, we ultimately demonstrate the potential for integrating domain-specific accelerators with the next generation of neural networks., Comment: 10 pages, 9 figures, journal

Spiking neural networks, also often referred to as the third generation of neural networks, carry the potential for a massive reduction in memory and energy consumption over traditional, second-generation neural networks. Inspired by the undisputed efficiency of the human brain, they introduce temporal and neuronal sparsity, which can be exploited by next-generation neuromorphic hardware. To open the pathway toward engineering applications, we introduce this exciting technology in the context of continuum mechanics. However, the nature of spiking neural networks poses a challenge for regression problems, which frequently arise in the modeling of engineering sciences. To overcome this problem, a framework for regression using spiking neural networks is proposed. In particular, a network topology for decoding binary spike trains to real numbers is introduced, utilizing the membrane potential of spiking neurons. As the aim of this contribution is a concise introduction to this new methodology, several different spiking neural architectures, ranging from simple spiking feed-forward to complex spiking long short-term memory neural networks, are derived. Several numerical experiments directed towards regression of linear and nonlinear, history-dependent material models are carried out. A direct comparison with counterparts of traditional neural networks shows that the proposed framework is much more efficient while retaining precision and generalizability. All code has been made publicly available in the interest of reproducibility and to promote continued enhancement in this new domain.

In-memory computing (IMC) systems have great potential for accelerating data-intensive tasks such as deep neural networks (DNNs). As DNN models are generally highly proprietary, the neural network architectures become valuable targets for attacks. In IMC systems, since the whole model is mapped on chip and weight memory read can be restricted, the pre-mapped DNN model acts as a ``black box'' for users. However, the localized and stationary weight and data patterns may subject IMC systems to other attacks. In this paper, we propose a side-channel attack methodology on IMC architectures. We show that it is possible to extract model architectural information from power trace measurements without any prior knowledge of the neural network. We first developed a simulation framework that can emulate the dynamic power traces of the IMC macros. We then performed side-channel leakage analysis to reverse engineer model information such as the stored layer type, layer sequence, output channel/feature size and convolution kernel size from power traces of the IMC macros. Based on the extracted information, full networks can potentially be reconstructed without any knowledge of the neural network. Finally, we discuss potential countermeasures for building IMC systems that offer resistance to these model extraction attack.

We present a fully memristive spiking neural network (MSNN) consisting of physically-realizable memristive neurons and memristive synapses to implement an unsupervised Spiking Time Dependent Plasticity (STDP) learning rule. The system is fully memristive in that both neuronal and synaptic dynamics can be realized by using memristors. The neuron is implemented using the SPICE-level memristive integrate-and-fire (MIF) model, which consists of a minimal number of circuit elements necessary to achieve distinct depolarization, hyperpolarization, and repolarization voltage waveforms. The proposed MSNN uniquely implements STDP learning by using cumulative weight changes in memristive synapses from the voltage waveform changes across the synapses, which arise from the presynaptic and postsynaptic spiking voltage signals during the training process. Two types of MSNN architectures are investigated: 1) a biologically plausible memory retrieval system, and 2) a multi-class classification system. Our circuit simulation results verify the MSNN's unsupervised learning efficacy by replicating biological memory retrieval mechanisms, and achieving 97.5% accuracy in a 4-pattern recognition problem in a large scale discriminative MSNN.

We present a fully memristive spiking neural network (MSNN) consisting of novel memristive neurons trained using the backpropagation through time (BPTT) learning rule. Gradient descent is applied directly to the memristive integrated-and-fire (MIF) neuron designed using analog SPICE circuit models, which generates distinct depolarization, hyperpolarization, and repolarization voltage waveforms. Synaptic weights are trained by BPTT using the membrane potential of the MIF neuron model and can be processed on memristive crossbars. The natural spiking dynamics of the MIF neuron model are fully differentiable, eliminating the need for gradient approximations that are prevalent in the spiking neural network literature. Despite the added complexity of training directly on SPICE circuit models, we achieve 97.58% accuracy on the MNIST testing dataset and 75.26% on the Fashion-MNIST testing dataset, the highest accuracies among all fully MSNNs.

Spiking and Quantized Neural Networks (NNs) are becoming exceedingly important for hyper-efficient implementations of Deep Learning (DL) algorithms. However, these networks face challenges when trained using error backpropagation, due to the absence of gradient signals when applying hard thresholds. The broadly accepted trick to overcoming this is through the use of biased gradient estimators: surrogate gradients which approximate thresholding in Spiking Neural Networks (SNNs), and Straight-Through Estimators (STEs), which completely bypass thresholding in Quantized Neural Networks (QNNs). While noisy gradient feedback has enabled reasonable performance on simple supervised learning tasks, it is thought that such noise increases the difficulty of finding optima in loss landscapes, especially during the later stages of optimization. By periodically boosting the Learning Rate (LR) during training, we expect the network can navigate unexplored solution spaces that would otherwise be difficult to reach due to local minima, barriers, or flat surfaces. This paper presents a systematic evaluation of a cosine-annealed LR schedule coupled with weight-independent adaptive moment estimation as applied to Quantized SNNs (QSNNs). We provide a rigorous empirical evaluation of this technique on high precision and 4-bit quantized SNNs across three datasets, demonstrating (close to) state-of-the-art performance on the more complex datasets. Our source code is available at this link: https://github.com/jeshraghian/QSNNs.

Spiking neural networks can compensate for quantization error by encoding information either in the temporal domain, or by processing discretized quantities in hidden states of higher precision. In theory, a wide dynamic range state-space enables multiple binarized inputs to be accumulated together, thus improving the representational capacity of individual neurons. This may be achieved by increasing the firing threshold, but make it too high and sparse spike activity turns into no spike emission. In this paper, we propose the use of `threshold annealing' as a warm-up method for firing thresholds. We show it enables the propagation of spikes across multiple layers where neurons would otherwise cease to fire, and in doing so, achieve highly competitive results on four diverse datasets, despite using binarized weights. Source code is available at https://github.com/jeshraghian/snn-tha

The impact of device and circuit-level effects in mixed-signal Resistive Random Access Memory (RRAM) accelerators typically manifest as performance degradation of Deep Learning (DL) algorithms, but the degree of impact varies based on algorithmic features. These include network architecture, capacity, weight distribution, and the type of inter-layer connections. Techniques are continuously emerging to efficiently train sparse neural networks, which may have activation sparsity, quantization, and memristive noise. In this paper, we present an extended Design Space Exploration (DSE) methodology to quantify the benefits and limitations of dense and sparse mapping schemes for a variety of network architectures. While sparsity of connectivity promotes less power consumption and is often optimized for extracting localized features, its performance on tiled RRAM arrays may be more susceptible to noise due to under-parameterization, when compared to dense mapping schemes. Moreover, we present a case study quantifying and formalizing the trade-offs of typical non-idealities introduced into 1-Transistor-1-Resistor (1T1R) tiled memristive architectures and the size of modular crossbar tiles using the CIFAR-10 dataset., Comment: Accepted at 2022 IEEE International Symposium on Circuits and Systems (ISCAS). [v2] Fixed incorrectly labeled author affiliations for Chenqi Li, Amirali Amirsoleimani, and Roman Genov

We present MEMprop, the adoption of gradient-based learning to train fully memristive spiking neural networks (MSNNs). Our approach harnesses intrinsic device dynamics to trigger naturally arising voltage spikes. These spikes emitted by memristive dynamics are analog in nature, and thus fully differentiable, which eliminates the need for surrogate gradient methods that are prevalent in the spiking neural network (SNN) literature. Memristive neural networks typically either integrate memristors as synapses that map offline-trained networks, or otherwise rely on associative learning mechanisms to train networks of memristive neurons. We instead apply the backpropagation through time (BPTT) training algorithm directly on analog SPICE models of memristive neurons and synapses. Our implementation is fully memristive, in that synaptic weights and spiking neurons are both integrated on resistive RAM (RRAM) arrays without the need for additional circuits to implement spiking dynamics, e.g., analog-to-digital converters (ADCs) or thresholded comparators. As a result, higher-order electrophysical effects are fully exploited to use the state-driven dynamics of memristive neurons at run time. By moving towards non-approximate gradient-based learning, we obtain highly competitive accuracy amongst previously reported lightweight dense fully MSNNs on several benchmarks.

Spiking neural networks have shown much promise as an energy-efficient alternative to artificial neural networks. However, understanding the impacts of sensor noises and input encodings on the network activity and performance remains difficult with common neuromorphic vision baselines like classification. Therefore, we propose a spiking neural network approach for single object localization trained using surrogate gradient descent, for frame- and event-based sensors. We compare our method with similar artificial neural networks and show that our model has competitive/better performance in accuracy, robustness against various corruptions, and has lower energy consumption. Moreover, we study the impact of neural coding schemes for static images in accuracy, robustness, and energy efficiency. Our observations differ importantly from previous studies on bio-plausible learning rules, which helps in the design of surrogate gradient trained architectures, and offers insight to design priorities in future neuromorphic technologies in terms of noise characteristics and data encoding methods., Comment: 21 pages, 12 figures

Neural processor development is reducing our reliance on remote server access to process deep learning operations in an increasingly edge-driven world. By employing in-memory processing, parallelization techniques, and algorithm-hardware co-design, memristor crossbar arrays are known to efficiently compute large scale matrix-vector multiplications. However, state-of-the-art implementations of negative weights require duplicative column wires, and high precision weights using single-bit memristors further distributes computations. These constraints dramatically increase chip area and resistive losses, which lead to increased power consumption and reduced accuracy. In this paper, we develop an adaptive precision method by varying the number of memristors at each crosspoint. We also present a weight mapping algorithm designed for implementation on our crossbar array. This novel algorithm-hardware solution is described as the radix-X Convolutional Neural Network Crossbar Array, and demonstrate how to efficiently represent negative weights using a single column line, rather than double the number of additional columns. Using both simulation and experimental results, we verify that our radix-5 CNN array achieves a validation accuracy of 90.5% on the CIFAR-10 dataset, a 4.5% improvement over binarized neural networks whilst simultaneously reducing crossbar area by 46% over conventional arrays by removing the need for duplicate columns to represent signed weights., Comment: 12 pages, 17 figures

The 69th World Health Assembly endorsed the Global Health Sector Strategy for Viral Hepatitis, embracing a goal to eliminate hepatitis infection as a public health threat by 2030. This was followed by the World Health Organization's (WHO) global targets for the care and management of hepatitis B virus (HBV) and hepatitis C virus (HCV) infections. These announcements and targets were important in raising awareness and calling for action; however, tracking countries' progress towards these elimination goals has provided insights to the limitations of these targets. The existing targets compare a country's progress relative to its 2015 values, penalizing countries who started their programmes prior to 2015, countries with a young population, or countries with a low prevalence. We recommend that (1) WHO simplify the hepatitis elimination targets, (2) change to absolute targets and (3) allow countries to achieve these disease targets with their own service coverage initiatives that will have the maximum impact. The recommended targets are as follows: reduce HCV new chronic cases to <=5 per 100 000, reduce HBV prevalence among 1-year-olds to <=0.1%, reduce HBV and HCV mortality to <=5 per 100 000, and demonstrate HBV and HCV year-to-year decrease in new HCV- and HBV-related HCC cases. The objective of our recommendations is not to lower expectations or diminish the hepatitis elimination standards, but to provide clearer targets that recognize the past and current elimination efforts by countries, help measure progress towards true elimination, and motivate other countries to follow suit.Copyright © 2020 John Wiley & Sons Ltd

The 69th World Health Assembly endorsed the Global Health Sector Strategy for Viral Hepatitis, embracing a goal to eliminate hepatitis infection as a public health threat by 2030. This was followed by the World Health Organization's (WHO) global targets for the care and management of hepatitis B virus (HBV) and hepatitis C virus (HCV) infections. These announcements and targets were important in raising awareness and calling for action; however, tracking countries' progress towards these elimination goals has provided insights to the limitations of these targets. The existing targets compare a country's progress relative to its 2015 values, penalizing countries who started their programmes prior to 2015, countries with a young population, or countries with a low prevalence. We recommend that (1) WHO simplify the hepatitis elimination targets, (2) change to absolute targets and (3) allow countries to achieve these disease targets with their own service coverage initiatives that will have the maximum impact. The recommended targets are as follows: reduce HCV new chronic cases to <=5 per 100 000, reduce HBV prevalence among 1-year-olds to <=0.1%, reduce HBV and HCV mortality to <=5 per 100 000, and demonstrate HBV and HCV year-to-year decrease in new HCV- and HBV-related HCC cases. The objective of our recommendations is not to lower expectations or diminish the hepatitis elimination standards, but to provide clearer targets that recognize the past and current elimination efforts by countries, help measure progress towards true elimination, and motivate other countries to follow suit.Copyright © 2020 John Wiley & Sons Ltd

Wireless sensor nodes are heavily resource-constrained due to their edge form factor, which has motivated increasing battery life through low-power techniques. This paper proposes a power management method that leads to less energy consumption in an idle state than conventional power management systems used in wireless sensor nodes. We analyze and benchmark the power consumption between Sleep, Idle, and Run modes. To reduce sensor node power consumption, we develop fine-grained power modes (FGPM) with five states which modulate energy consumption according to the sensor node's communication status. We evaluate the proposed method on a test bench Mica2. As a result, the power consumed is 74.2% lower than that of conventional approaches. The proposed method targets the reduction of power consumption in IoT sensor modules with long sleep mode or short packet data in which most networks operate.

Background & Aims: Chronic liver disease (CLD) and cirrhosis are associated with immune dysregulation, leading to concerns that affected patients may be at risk of adverse outcomes following SARS-CoV-2 infection. We aimed to determine the impact of COVID-19 on patients with pre-existing liver disease, which currently remains ill-defined. Methods: Between 25th March and 8th July 2020, data on 745 patients with CLD and SARS-CoV-2 (including 386 with and 359 without cirrhosis) were collected by 2 international registries and compared to data on non-CLD patients with SARS-CoV-2 from a UK hospital network. Results: Mortality was 32% in patients with cirrhosis compared to 8% in those without (p <0.001). Mortality in patients with cirrhosis increased according to Child-Pugh class (A [19%], B [35%], C [51%]) and the main cause of death was from respiratory failure (71%). After adjusting for baseline characteristics, factors associated with death in the total CLD cohort were age (odds ratio [OR] 1.02; 1.01–1.04), Child-Pugh A (OR 1.90; 1.03–3.52), B (OR 4.14; 2.4–7.65), or C (OR 9.32; 4.80–18.08) cirrhosis and alcohol-related liver disease (OR 1.79; 1.03–3.13). Compared to patients without CLD (n = 620), propensity-score-matched analysis revealed significant increases in mortality in those with Child-Pugh B (+20.0% [8.8%–31.3%]) and C (+38.1% [27.1%–49.2%]) cirrhosis. Acute hepatic decompensation occurred in 46% of patients with cirrhosis, of whom 21% had no respiratory symptoms. Half of those with hepatic decompensation had acute-on-chronic liver failure. Conclusions: In the largest such cohort to date, we demonstrate that baseline liver disease stage and alcohol-related liver disease are independent risk factors for death from COVID-19. These data have important implications for the risk stratification of patients with CLD across the globe during the COVID-19 pandemic. Lay summary: This international registry study demonstrates that patients with cirrhosis are at increased

We invited authors of selected Comments and Perspectives published in Nature Machine Intelligence in the latter half of 2019 and first half of 2020 to describe how their topic has developed, what their thoughts are about the challenges of 2020, and what they look forward to in 2021., Postprint (author's final draft)

The brain is the perfect place to look for inspiration to develop more efficient neural networks. The inner workings of our synapses and neurons provide a glimpse at what the future of deep learning might look like. This paper serves as a tutorial and perspective showing how to apply the lessons learnt from several decades of research in deep learning, gradient descent, backpropagation and neuroscience to biologically plausible spiking neural neural networks. We also explore the delicate interplay between encoding data as spikes and the learning process; the challenges and solutions of applying gradient-based learning to spiking neural networks (SNNs); the subtle link between temporal backpropagation and spike timing dependent plasticity, and how deep learning might move towards biologically plausible online learning. Some ideas are well accepted and commonly used amongst the neuromorphic engineering community, while others are presented or justified for the first time here. The fields of deep learning and spiking neural networks evolve very rapidly. We endeavour to treat this document as a 'dynamic' manuscript that will continue to be updated as the common practices in training SNNs also change. A series of companion interactive tutorials complementary to this paper using our Python package, snnTorch, are also made available. See https://snntorch.readthedocs.io/en/latest/tutorials/index.html .

The problem of worst case edge deletion from a network is considered. Suppose that you have a communication network and you can delete a single edge. Which edge deletion causes the largest disruption? More formally, given a graph, which edge after deletion disconnects the maximum number of pairs of vertices, where ties for number of pairs disconnected are broken by finding an edge that increases the average shortest path length the maximum amount. This problem is interesting both practically and theoretically. We call it the \emph{single edge deletion problem}. Our contributions include formally defining the single edge deletion problem and providing motivations from network analysis. Also, we give an algorithm that solves the problem much faster than a naive solution. The algorithm incorporates sophisticated and novel techniques, and generalises to the problem of computing the all-pairs shortest paths table after deleting each edge individually. This means the algorithm has deep theoretical interest as well as the potential for even wider applications than those we present here., Comment: Found previous published results with similar findings

The search for a compatible application of memristor-CMOS logic gates has remained elusive, as the data density benefits are offset by slow switching speeds and resistive dissipation. Active microdisplays typically prioritize pixel density (and therefore resolution) over that of speed, where the most widely used refresh rates fall between 25-240 Hz. Therefore, memristor-CMOS logic is a promising fit for peripheral IO logic in active matrix displays. In this paper, we design and implement a ternary 1-3 line decoder and a ternary 2-9 line decoder which are used to program a seven segment LED display. SPICE simulations are conducted in a 50-nm process, and the decoders are synthesized on an Altera Cyclone IV field-programmable gate array (FPGA) development board which implements a ternary memristor model designed in Quartus II. We compare our hardware results to a binary coded decimal (BCD)-to-seven segment display decoder, and show our memristor-CMOS approach reduces the total IO power consumption by a factor of approximately 6 times at a maximum synthesizable frequency of 293.77MHz. Although the speed is approximately half of the native built-in BCD-to-seven decoder, the comparatively slow refresh rates of typical microdisplays indicate this to be a tolerable trade-off, which promotes data density over speed.

Stochastic Computing (SC) is a computing paradigm that allows for the low-cost and low-power computation of various arithmetic operations using stochastic bit streams and digital logic. In contrast to conventional representation schemes used within the binary domain, the sequence of bit streams in the stochastic domain is inconsequential, and computation is usually non-deterministic. In this brief, we exploit the stochasticity during switching of probabilistic Conductive Bridging RAM (CBRAM) devices to efficiently generate stochastic bit streams in order to perform Deep Learning (DL) parameter optimization, reducing the size of Multiply and Accumulate (MAC) units by 5 orders of magnitude. We demonstrate that in using a 40-nm Complementary Metal Oxide Semiconductor (CMOS) process our scalable architecture occupies 1.55mm$^2$ and consumes approximately 167$\mu$W when optimizing parameters of a Convolutional Neural Network (CNN) while it is being trained for a character recognition task, observing no notable reduction in accuracy post-training., Comment: Accepted by IEEE Transactions on Circuits and Systems Part II: Express Briefs

Dynamic vision sensors are able to operate at high temporal resolutions within resource constrained environments, though at the expense of capturing static content. The sparse nature of event streams enables efficient downstream processing tasks as they are suited for power-efficient spiking neural networks. One of the challenges associated with neuromorphic vision is the lack of interpretability of event streams. While most application use-cases do not intend for the event stream to be visually interpreted by anything other than a classification network, there is a lost opportunity to integrating these sensors in spaces that conventional high-speed CMOS sensors cannot go. For example, biologically invasive sensors such as endoscopes must fit within stringent power budgets, which do not allow MHz-speeds of image integration. While dynamic vision sensing can fill this void, the interpretation challenge remains and will degrade confidence in clinical diagnostics. The use of generative adversarial networks presents a possible solution to overcoming and compensating for a vision chip's poor spatial resolution and lack of interpretability. In this paper, we methodically apply the Pix2Pix network to naturalize the event stream from spike-converted CIFAR-10 and Linnaeus 5 datasets. The quality of the network is benchmarked by performing image classification of naturalized event streams, which converges to within 2.81% of equivalent raw images, and an associated improvement over unprocessed event streams by 13.19% for the CIFAR-10 and Linnaeus 5 datasets., Comment: 5 pages, 7 figures

Deep neural network inference accelerators are rapidly growing in importance as we turn to massively parallelized processing beyond GPUs and ASICs. The dominant operation in feedforward inference is the multiply-and-accumlate process, where each column in a crossbar generates the current response of a single neuron. As a result, memristor crossbar arrays parallelize inference and image processing tasks very efficiently. In this brief, we present a 3-D active memristor crossbar array `CrossStack', which adopts stacked pairs of Al/TiO2/TiO2-x/Al devices with common middle electrodes. By designing CMOS-memristor hybrid cells used in the layout of the array, CrossStack can operate in one of two user-configurable modes as a reconfigurable inference engine: 1) expansion mode and 2) deep-net mode. In expansion mode, the resolution of the network is doubled by increasing the number of inputs for a given chip area, reducing IR drop by 22%. In deep-net mode, inference speed per-10-bit convolution is improved by 29\% by simultaneously using one TiO2/TiO2-x layer for read processes, and the other for write processes. We experimentally verify both modes on our $10\times10\times2$ array., Comment: 5 pages, 4 figures

The advent of dedicated Deep Learning (DL) accelerators and neuromorphic processors has brought on new opportunities for applying both Deep and Spiking Neural Network (SNN) algorithms to healthcare and biomedical applications at the edge. This can facilitate the advancement of medical Internet of Things (IoT) systems and Point of Care (PoC) devices. In this paper, we provide a tutorial describing how various technologies including emerging memristive devices, Field Programmable Gate Arrays (FPGAs), and Complementary Metal Oxide Semiconductor (CMOS) can be used to develop efficient DL accelerators to solve a wide variety of diagnostic, pattern recognition, and signal processing problems in healthcare. Furthermore, we explore how spiking neuromorphic processors can complement their DL counterparts for processing biomedical signals. The tutorial is augmented with case studies of the vast literature on neural network and neuromorphic hardware as applied to the healthcare domain. We benchmark various hardware platforms by performing a sensor fusion signal processing task combining electromyography (EMG) signals with computer vision. Comparisons are made between dedicated neuromorphic processors and embedded AI accelerators in terms of inference latency and energy. Finally, we provide our analysis of the field and share a perspective on the advantages, disadvantages, challenges, and opportunities that various accelerators and neuromorphic processors introduce to healthcare and biomedical domains., Comment: Accepted by IEEE Transactions on Biomedical Circuits and Systems (21 pages, 10 figures, 5 tables)

The advent of dedicated Deep Learning (DL) accelerators and neuromorphic processors has brought on new opportunities for applying both Deep and Spiking Neural Network (SNN) algorithms to healthcare and biomedical applications at the edge. This can facilitate the advancement of medical Internet of Things (IoT) systems and Point of Care (PoC) devices. In this paper, we provide a tutorial describing how various technologies including emerging memristive devices, Field Programmable Gate Arrays (FPGAs), and Complementary Metal Oxide Semiconductor (CMOS) can be used to develop efficient DL accelerators to solve a wide variety of diagnostic, pattern recognition, and signal processing problems in healthcare. Furthermore, we explore how spiking neuromorphic processors can complement their DL counterparts for processing biomedical signals. The tutorial is augmented with case studies of the vast literature on neural network and neuromorphic hardware as applied to the healthcare domain. We benchmark various hardware platforms by performing a sensor fusion signal processing task combining electromyography (EMG) signals with computer vision. Comparisons are made between dedicated neuromorphic processors and embedded AI accelerators in terms of inference latency and energy. Finally, we provide our analysis of the field and share a perspective on the advantages, disadvantages, challenges, and opportunities that various accelerators and neuromorphic processors introduce to healthcare and biomedical domains.

Background: Epilepsy, which develops in the elderly, is recognized as a major health burden. Although health-related quality of life (HRQoL) is an essential element in the medical treatment of elderly patients with epilepsy, it is a question whether epilepsy and its treatment effectively influence the quality of life (QoL) in the elderly. Objectives: The current study aimed at evaluating the relationship between demographic and clinical aspects of epilepsy in HRQoL of elderly patients. Since HRQoL scores are bounded, the Kumaraswamy (Kum) regression model was used to analyze the data. Methods: The current study was conducted on 766 elderly patients diagnosed with epilepsy taking at least one antiepileptic drug (AED) selected from six neurologic clinics in Iran. In addition to demographic information, the Liverpool seizure severity scale (LSSS), medication adherence report scale (MARS-5), and quality of life in epilepsy (QoLIE-31) questionnaire were completed for patients. Data were analyzed using multiple linear regression (MLR) and the Kum regression models. Results: Most of the patients included in the study had focal (70.2%) epilepsy. Mean duration of disease was 17.71 +/- 4.56 years and the average number of seizures was 3.4 +/- 3.2 episodes per month. The Kum regression model indicated that seizure frequency (beta = 0.157, P < 0.0001) and LSSS score (beta = -0.003, P = 0.009) were significant and negative predictors of overall QoLEI-31 score; MARS-5 score was a positive predictor of overall QoLEI-31 score (beta = 0.014, P= 0.002). However, disease duration and serum AED level had no significant effects on overall QoLEI-31 score. Conclusions: The findings suggested that increased seizure frequencyand severity were associated with lower QoL and medication adherence was directly associated with HRQoL. The Kum regression could be a suitable alternative to the methods currently used in the analysis of HRQoL data.

The Australian Commonwealth government recently announced a grant of $4.75 million as part of a $13.5 million program to establish a world class networked IC tele-test facility in Australia. The facility will be based on a state-of-the-art semiconductor tester located at Edith Cowan University in Perth that will operate as a virtual centre spanning Australia. Satellite nodes will be located at the University of Western Australia, Griffith University, Macquarie University, Victoria University and the University of Adelaide. The facility will provide vital equipment to take Australia to the frontier of critically important and expanding fields in microelectronics research and development. The tele-test network will provide state of the art environment for the electronics and microelectronics research and the industry community around Australia to test and prototype Very Large Scale Integrated (VLSI) circuits and other System On a Chip (SOC) devices, prior to moving to the manufacturing stage. Such testing is absolutely essential to ensure that the device performs to specification. This paper presents the current context in which the testing facility is being established, the methodologies behind the integration of design and test strategies and the target shape of the tele-testing Facility.