Ferroelectric capacitors and field-effect transistors as in-memory computing elements for machine learning workloads

This study discusses the feasibility of Ferroelectric Capacitors (FeCaps) and Ferroelectric Field-Effect Transistors (FeFETs) as In-Memory Computing (IMC) elements to accelerate machine learning (ML) workloads. We conducted an exploration of device fabrication and proposed system-algorithm co-design to boost performance. A novel FeCap device, incorporating an interfacial layer (IL) andHf 0.5 Zr 0.5 O 2 (HZO), ensures a reduction in operating voltage and enhances HZO scaling while being compatible with CMOS circuits. The IL also enriches ferroelectricity and retention properties. When integrated into crossbar arrays, FeCaps and FeFETs demonstrate their effectiveness as IMC components, eliminating sneak paths and enabling selector-less operation, leading to notable improvements in energy efficiency and area utilization. However, it is worth noting that limited capacitance ratios in FeCaps introduced errors in multiply-and-accumulate (MAC) computations. The proposed co-design approach helps in mitigating these errors and achieves high accuracy in classifying the CIFAR-10 dataset, elevating it from a baseline of 10% to 81.7%. FeFETs in crossbars, with a higher on-off ratio, outperform FeCaps, and our proposed charge-based sensing scheme achieved at least an order of magnitude reduction in power consumption, compared to prevalent current-based methods.

First report on pesticide sub-chronic and chronic toxicities against dogs using QSAR and chemical read-across

Excessive use of chemicals is the outcome of the industrialization of agricultural sectors which leads to disturbance of ecological balance. Various agrochemicals are widely used in agricultural fields, urban green areas, and to protect from various pest-associated diseases. Due to their long-term health and environmental hazards, chronic toxicity assessment is crucial. Since in vivo and in vitro toxicity assessments are costly, lengthy, and require a large number of animal experiments, in silico toxicity approaches are better alternatives to save time, cost, and animal experimentation. We have developed the first regression-based 2D-QSAR models using different sub-chronic and chronic toxicity data of pesticides against dogs employing 2D descriptors. From the statistical results (n train = 53 - 62 ,   r 2  = 0.614 to 0.754, Q L O O 2  = 0.501 to 0.703 and   Q F 1 2  = 0.531 to 0.718, Q F 2 2 = 0.523 - 0.713 ), it was concluded that the models are robust, reliable, interpretable, and predictive. Similarity-based read-across algorithm was also used to improve the predictivity (Q F 1 2 = 0.595 - 0.813 , Q F 2 2 = 0.573 - 0.809 ) of the models. 5132 chemicals obtained from the CPDat and 1694 pesticides obtained from the PPDB database were also screened using the developed models, and their predictivity and reliability were checked. Thus, these models will be helpful for eco-toxicological data-gap filling, toxicity prediction of untested pesticides, and development of novel, safer & eco-friendly pesticides.

RoMIA: a framework for creating Robust Medical Imaging AI models for chest radiographs

Artificial Intelligence (AI) methods, particularly Deep Neural Networks (DNNs), have shown great promise in a range of medical imaging tasks. However, the susceptibility of DNNs to producing erroneous outputs under the presence of input noise and variations is of great concern and one of the largest challenges to their adoption in medical settings. Towards addressing this challenge, we explore the robustness of DNNs trained for chest radiograph classification under a range of perturbations reflective of clinical settings. We propose RoMIA, a framework for the creation of Robust Medical Imaging AI models. RoMIA adds three key steps to the model training and deployment flow: (i) Noise-added training, wherein a part of the training data is synthetically transformed to represent common noise sources, (ii) Fine-tuning with input mixing, in which the model is refined with inputs formed by mixing data from the original training set with a small number of images from a different source, and (iii) DCT-based denoising, which removes a fraction of high-frequency components of each image before applying the model to classify it. We applied RoMIA to create six different robust models for classifying chest radiographs using the CheXpert dataset. We evaluated the models on the CheXphoto dataset, which consists of naturally and synthetically perturbed images intended to evaluate robustness. Models produced by RoMIA show 3%-5% improvement in robust accuracy, which corresponds to an average reduction of 22.6% in misclassifications. These results suggest that RoMIA can be a useful step towards enabling the adoption of AI models in medical imaging applications.

CL3: Generalization of Contrastive Loss for Lifelong Learning

Lifelong learning portrays learning gradually in nonstationary environments and emulates the process of human learning, which is efficient, robust, and able to learn new concepts incrementally from sequential experience. To equip neural networks with such a capability, one needs to overcome the problem of catastrophic forgetting, the phenomenon of forgetting past knowledge while learning new concepts. In this work, we propose a novel knowledge distillation algorithm that makes use of contrastive learning to help a neural network to preserve its past knowledge while learning from a series of tasks. Our proposed generalized form of contrastive distillation strategy tackles catastrophic forgetting of old knowledge, and minimizes semantic drift by maintaining a similar embedding space, as well as ensures compactness in feature distribution to accommodate novel tasks in a current model. Our comprehensive study shows that our method achieves improved performances in the challenging class-incremental, task-incremental, and domain-incremental learning for supervised scenarios.

QSAR assessment of aquatic toxicity potential of diverse agrochemicals

The fast-increasing number of commercially produced chemicals challenges the experimental ecotoxicity assessment methods, which are costly, time-consuming, and dependent on the sacrifice of animals. In this regard, Quantitative Structure-Property/Activity Relationships (QSPR/QSAR) have led the way in developing ecotoxicity assessment models. In this study, QSAR models have been developed using the pEC50 values of 82 diverse agrochemicals or agro-molecules against a planktonic crustacean Daphnia magna with easily interpretable 2D descriptors. Moreover, a link among octanol-water partition coefficient (KOW), bio-concentration factor (BCF), and critical body residue (CBR) has been addressed, and their imputation for the prediction of the toxicity endpoint (EC50) has been done with an objective of the advanced exploration of several ecotoxicological parameters for toxic chemicals. The developed partial least squares (PLS) models were validated rigorously and proved to be robust, sound, and immensely well-predictive. The final Daphnia toxicity model derived from experimental derived properties along with computed descriptors emerged better in statistical quality and predictivity than those obtained solely from computed descriptors. Additionally, the pEC50 and other important properties (log KOW, log BCF, and log CBR) for a set of external agro-molecules, not employed in model development, were predicted to show the predictive ability of the models.

Multiorgan MRI findings after hospitalisation with COVID-19 in the UK (C-MORE): a prospective, multicentre, observational cohort study

INTRODUCTION

The multiorgan impact of moderate to severe coronavirus infections in the post-acute phase is still poorly understood. We aimed to evaluate the excess burden of multiorgan abnormalities after hospitalisation with COVID-19, evaluate their determinants, and explore associations with patient-related outcome measures.

METHODS

In a prospective, UK-wide, multicentre MRI follow-up study (C-MORE), adults (aged ≥18 years) discharged from hospital following COVID-19 who were included in Tier 2 of the Post-hospitalisation COVID-19 study (PHOSP-COVID) and contemporary controls with no evidence of previous COVID-19 (SARS-CoV-2 nucleocapsid antibody negative) underwent multiorgan MRI (lungs, heart, brain, liver, and kidneys) with quantitative and qualitative assessment of images and clinical adjudication when relevant. Individuals with end-stage renal failure or contraindications to MRI were excluded. Participants also underwent detailed recording of symptoms, and physiological and biochemical tests. The primary outcome was the excess burden of multiorgan abnormalities (two or more organs) relative to controls, with further adjustments for potential confounders. The C-MORE study is ongoing and is registered with ClinicalTrials.gov, NCT04510025.

FINDINGS

Of 2710 participants in Tier 2 of PHOSP-COVID, 531 were recruited across 13 UK-wide C-MORE sites. After exclusions, 259 C-MORE patients (mean age 57 years [SD 12]; 158 [61%] male and 101 [39%] female) who were discharged from hospital with PCR-confirmed or clinically diagnosed COVID-19 between March 1, 2020, and Nov 1, 2021, and 52 non-COVID-19 controls from the community (mean age 49 years [SD 14]; 30 [58%] male and 22 [42%] female) were included in the analysis. Patients were assessed at a median of 5·0 months (IQR 4·2-6·3) after hospital discharge. Compared with non-COVID-19 controls, patients were older, living with more obesity, and had more comorbidities. Multiorgan abnormalities on MRI were more frequent in patients than in controls (157 [61%] of 259 vs 14 [27%] of 52; p<0·0001) and independently associated with COVID-19 status (odds ratio [OR] 2·9 [95% CI 1·5-5·8]; padjusted=0·0023) after adjusting for relevant confounders. Compared with controls, patients were more likely to have MRI evidence of lung abnormalities (p=0·0001; parenchymal abnormalities), brain abnormalities (p<0·0001; more white matter hyperintensities and regional brain volume reduction), and kidney abnormalities (p=0·014; lower medullary T1 and loss of corticomedullary differentiation), whereas cardiac and liver MRI abnormalities were similar between patients and controls. Patients with multiorgan abnormalities were older (difference in mean age 7 years [95% CI 4-10]; mean age of 59·8 years [SD 11·7] with multiorgan abnormalities vs mean age of 52·8 years [11·9] without multiorgan abnormalities; p<0·0001), more likely to have three or more comorbidities (OR 2·47 [1·32-4·82]; padjusted=0·0059), and more likely to have a more severe acute infection (acute CRP >5mg/L, OR 3·55 [1·23-11·88]; padjusted=0·025) than those without multiorgan abnormalities. Presence of lung MRI abnormalities was associated with a two-fold higher risk of chest tightness, and multiorgan MRI abnormalities were associated with severe and very severe persistent physical and mental health impairment (PHOSP-COVID symptom clusters) after hospitalisation.

INTERPRETATION

After hospitalisation for COVID-19, people are at risk of multiorgan abnormalities in the medium term. Our findings emphasise the need for proactive multidisciplinary care pathways, with the potential for imaging to guide surveillance frequency and therapeutic stratification.

FUNDING

UK Research and Innovation and National Institute for Health Research.

Subspace distillation for continual learning

An ultimate objective in continual learning is to preserve knowledge learned in preceding tasks while learning new tasks. To mitigate forgetting prior knowledge, we propose a novel knowledge distillation technique that takes into the account the manifold structure of the latent/output space of a neural network in learning novel tasks. To achieve this, we propose to approximate the data manifold up-to its first order, hence benefiting from linear subspaces to model the structure and maintain the knowledge of a neural network while learning novel concepts. We demonstrate that the modeling with subspaces provides several intriguing properties, including robustness to noise and therefore effective for mitigating Catastrophic Forgetting in continual learning. We also discuss and show how our proposed method can be adopted to address both classification and segmentation problems. Empirically, we observe that our proposed method outperforms various continual learning methods on several challenging datasets including Pascal VOC, and Tiny-Imagenet. Furthermore, we show how the proposed method can be seamlessly combined with existing learning approaches to improve their performances. The codes of this article will be available at https://github.com/csiro-robotics/SDCL.

Compute in-Memory with Non-Volatile Elements for Neural Networks: A Review from a Co-Design Perspective

Deep learning has become ubiquitous, touching daily lives across the globe. Today, traditional computer architectures are stressed to their limits in efficiently executing the growing complexity of data and models. Compute-in-memory (CIM) can potentially play an important role in developing efficient hardware solutions that reduce data movement from compute-unit to memory, known as the von Neumann bottleneck. At its heart is a cross-bar architecture with nodal non-volatile-memory elements that performs an analog multiply-and-accumulate operation, enabling the matrix-vector-multiplications repeatedly used in all neural network workloads. The memory materials can significantly influence final system-level characteristics and chip performance, including speed, power, and classification accuracy. With an over-arching co-design viewpoint, this review assesses the use of cross-bar based CIM for neural networks, connecting the material properties and the associated design constraints and demands to application, architecture, and performance. Both digital and analog memory are considered, assessing the status for training and inference, and providing metrics for the collective set of properties non-volatile memory materials will need to demonstrate for a successful CIM technology.

Understanding mechanism governing the inflammatory potential of metal oxide nanoparticles using periodic table-based descriptors: a nano-QSAR approach

Metal oxide nanoparticles (MeOxNPs) can be made safer by understanding the interaction between the immune system and nanoparticles. A nano-quantitative structure-activity relationship (nano-QSAR) model can be used to evaluate nanoparticle risk quickly and conveniently. The present work attempts to develop nano-QSAR models to determine the inflammatory potential of MeOxNPs based on the THP-1 cell line. A comprehensive dataset comprising 32 MeOxNPs was used to develop a regression model with fold change (FC) of pro-inflammatory cytokine interleukin (IL)-1beta (IL-1b) release in the THP-1 cell line as the endpoint. Further, the same number of MeOx NPs with varying doses was modelled for the cell viability [-ln(p/(1-p))] endpoint. The results obtained from regression models were statistically significant. The descriptors obtained from the developed predictive models inferred that dose, electronegativity and the presence of metal ions and oxygen can be responsible for IL-1β leakage from the THP-1 cell line. Based on our results, we can conclude that periodic table-based descriptors, incorporated into the QSAR models, are reliable for modelling pro-inflammatory potential. Researchers can use these comprehensive results to design metal oxide nanoparticles with lower toxicity and determine the cause of pro-inflammatory conditions induced by MeOxNPs.

DIET-SNN: A Low-Latency Spiking Neural Network With Direct Input Encoding and Leakage and Threshold Optimization

Bioinspired spiking neural networks (SNNs), operating with asynchronous binary signals (or spikes) distributed over time, can potentially lead to greater computational efficiency on event-driven hardware. The state-of-the-art SNNs suffer from high inference latency, resulting from inefficient input encoding and suboptimal settings of the neuron parameters (firing threshold and membrane leak). We propose DIET-SNN, a low-latency deep spiking network trained with gradient descent to optimize the membrane leak and the firing threshold along with other network parameters (weights). The membrane leak and threshold of each layer are optimized with end-to-end backpropagation to achieve competitive accuracy at reduced latency. The input layer directly processes the analog pixel values of an image without converting it to spike train. The first convolutional layer converts analog inputs into spikes where leaky-integrate-and-fire (LIF) neurons integrate the weighted inputs and generate an output spike when the membrane potential crosses the trained firing threshold. The trained membrane leak selectively attenuates the membrane potential, which increases activation sparsity in the network. The reduced latency combined with high activation sparsity provides massive improvements in computational efficiency. We evaluate DIET-SNN on image classification tasks from CIFAR and ImageNet datasets on VGG and ResNet architectures. We achieve top-1 accuracy of 69% with five timesteps (inference latency) on the ImageNet dataset with 12× less compute energy than an equivalent standard artificial neural network (ANN). In addition, DIET-SNN performs 20- 500× faster inference compared to other state-of-the-art SNN models.

Prediction of soil ecotoxicity against Folsomia candida using acute and chronic endpoints

Soil invertebrates serve as great biological indicators of soil quality. However, there are very few in silico models developed so far on the soil toxicity of chemicals against soil invertebrates due to paucity of data. In this study, three available soil ecotoxicity data (pLC₅₀, pLOEL and pNOEL) against the soil invertebrate Folsomia candida were collected from the ECOTOX database (cfpub.epa.gov/ecotox) and subjected to quantitative structure-activity relationship (QSAR) analysis using 2D descriptors. The collected data for each endpoint were initially curated and used to develop a partial least squares (PLS) regression model based on the features selected through a genetic algorithm followed by the best subset selection. Both internal and external validation metrics of the models' predictions are well-balanced and within the acceptable range as per the Organization for the Economic Cooperation and Development (OECD) criteria. From the developed models, it has been found that molecular weight and presence of phosphate group, electron donor groups, and polyhalogen substitution have a significant impact on the soil ecotoxicity. The soil ecotoxicological risk assessment of organic chemicals can therefore be prioritized by these features. With the availability of additional data in the future, the models may be further refined for more precise predictions.

A machine learning pipeline to classify foetal heart rate deceleration with optimal feature set

Deceleration is considered a commonly practised means to assess Foetal Heart Rate (FHR) through visual inspection and interpretation of patterns in Cardiotocography (CTG). The precision of deceleration classification relies on the accurate estimation of corresponding event points (EP) from the FHR and the Uterine Contraction Pressure (UCP). This work proposes a deceleration classification pipeline by comparing four machine learning (ML) models, namely, Multilayer Perceptron (MLP), Random Forest (RF), Naïve Bayes (NB), and Simple Logistics Regression. Towards an automated classification of deceleration from EP using the pipeline, it systematically compares three approaches to create feature sets from the detected EP: (1) a novel fuzzy logic (FL)-based approach, (2) expert annotation by clinicians, and (3) calculated using National Institute of Child Health and Human Development guidelines. The classification results were validated using different popular statistical metrics, including receiver operating characteristic curve, intra-class correlation coefficient, Deming regression, and Bland-Altman Plot. The highest classification accuracy (97.94%) was obtained with MLP when the EP was annotated with the proposed FL approach compared to RF, which obtained 63.92% with the clinician-annotated EP. The results indicate that the FL annotated feature set is the optimal one for classifying deceleration from FHR.

ProKnow: Process knowledge for safety constrained and explainable question generation for mental health diagnostic assistance

Virtual Mental Health Assistants (VMHAs) are utilized in health care to provide patient services such as counseling and suggestive care. They are not used for patient diagnostic assistance because they cannot adhere to safety constraints and specialized clinical process knowledge (ProKnow) used to obtain clinical diagnoses. In this work, we define ProKnow as an ordered set of information that maps to evidence-based guidelines or categories of conceptual understanding to experts in a domain. We also introduce a new dataset of diagnostic conversations guided by safety constraints and ProKnow that healthcare professionals use (ProKnow-data). We develop a method for natural language question generation (NLG) that collects diagnostic information from the patient interactively (ProKnow-algo). We demonstrate the limitations of using state-of-the-art large-scale language models (LMs) on this dataset. ProKnow-algo incorporates the process knowledge through explicitly modeling safety, knowledge capture, and explainability. As computational metrics for evaluation do not directly translate to clinical settings, we involve expert clinicians in designing evaluation metrics that test four properties: safety, logical coherence, and knowledge capture for explainability while minimizing the standard cross entropy loss to preserve distribution semantics-based similarity to the ground truth. LMs with ProKnow-algo generated 89% safer questions in the depression and anxiety domain (tested property: safety). Further, without ProKnow-algo generations question did not adhere to clinical process knowledge in ProKnow-data (tested property: knowledge capture). In comparison, ProKnow-algo-based generations yield a 96% reduction in our metrics to measure knowledge capture. The explainability of the generated question is assessed by computing similarity with concepts in depression and anxiety knowledge bases. Overall, irrespective of the type of LMs, ProKnow-algo achieved an averaged 82% improvement over simple pre-trained LMs on safety, explainability, and process-guided question generation. For reproducibility, we will make ProKnow-data and the code repository of ProKnow-algo publicly available upon acceptance.

LWSNet - a novel deep-learning architecture to segregate Covid-19 and pneumonia from x-ray imagery

Automatic detection of lung diseases using AI-based tools became very much necessary to handle the huge number of cases occurring across the globe and support the doctors. This paper proposed a novel deep learning architecture named LWSNet (Light Weight Stacking Network) to separate Covid-19, cold pneumonia, and normal chest x-ray images. This framework is based on single, double, triple, and quadruple stack mechanisms to address the above-mentioned tri-class problem. In this framework, a truncated version of standard deep learning models and a lightweight CNN model was considered to conviniently deploy in resource-constraint devices. An evaluation was conducted on three publicly available datasets alongwith their combination. We received 97.28%, 96.50%, 97.41%, and 98.54% highest classification accuracies using quadruple stack. On further investigation, we found, using LWSNet, the average accuracy got improved from individual model to quadruple model by 2.31%, 2.55%, 2.88%, and 2.26% on four respective datasets.

LWSNet - a novel deep-learning architecture to segregate Covid-19 and pneumonia from x-ray imagery

Automatic detection of lung diseases using AI-based tools became very much necessary to handle the huge number of cases occurring across the globe and support the doctors. This paper proposed a novel deep learning architecture named LWSNet (Light Weight Stacking Network) to separate Covid-19, cold pneumonia, and normal chest x-ray images. This framework is based on single, double, triple, and quadruple stack mechanisms to address the above-mentioned tri-class problem. In this framework, a truncated version of standard deep learning models and a lightweight CNN model was considered to conviniently deploy in resource-constraint devices. An evaluation was conducted on three publicly available datasets alongwith their combination. We received 97.28%, 96.50%, 97.41%, and 98.54% highest classification accuracies using quadruple stack. On further investigation, we found, using LWSNet, the average accuracy got improved from individual model to quadruple model by 2.31%, 2.55%, 2.88%, and 2.26% on four respective datasets.

Application of Machine Learning and Deep Learning Techniques for COVID-19 Screening Using Radiological Imaging: A Comprehensive Review

Lung, being one of the most important organs in human body, is often affected by various SARS diseases, among which COVID-19 has been found to be the most fatal disease in recent times. In fact, SARS-COVID 19 led to pandemic that spreads fast among the community causing respiratory problems. Under such situation, radiological imaging-based screening [mostly chest X-ray and computer tomography (CT) modalities] has been performed for rapid screening of the disease as it is a non-invasive approach. Due to scarcity of physician/chest specialist/expert doctors, technology-enabled disease screening techniques have been developed by several researchers with the help of artificial intelligence and machine learning (AI/ML). It can be remarkably observed that the researchers have introduced several AI/ML/DL (deep learning) algorithms for computer-assisted detection of COVID-19 using chest X-ray and CT images. In this paper, a comprehensive review has been conducted to summarize the works related to applications of AI/ML/DL for diagnostic prediction of COVID-19, mainly using X-ray and CT images. Following the PRISMA guidelines, total 265 articles have been selected out of 1715 published articles till the third quarter of 2021. Furthermore, this review summarizes and compares varieties of ML/DL techniques, various datasets, and their results using X-ray and CT imaging. A detailed discussion has been made on the novelty of the published works, along with advantages and limitations.

Low precision decentralized distributed training over IID and non-IID data

Decentralized distributed learning is the key to enabling large-scale machine learning (training) on the edge devices utilizing private user-generated local data, without relying on the cloud. However, practical realization of such on-device training is limited by the communication and compute bottleneck. In this paper, we propose and show the convergence of low precision decentralized training that aims to reduce the computational complexity and communication cost of decentralized training. Many feedback-based compression techniques have been proposed in the literature to reduce communication costs. To the best of our knowledge, there is no work that applies and shows compute efficient training techniques such as quantization, pruning etc., for peer-to-peer decentralized learning setups. Since real-world applications have a significant skew in the data distribution, we design "Range-EvoNorm" as the normalization activation layer which is better suited for low precision training over non-IID data. Moreover, we show that the proposed low precision training can be used in synergy with other communication compression methods decreasing the communication cost further. Our experiments indicate that 8-bit decentralized training has minimal accuracy loss compared to its full precision counterpart even with non-IID data. However, when low precision training is accompanied by communication compression through sparsification we observe a 1-2% drop in accuracy. The proposed low precision decentralized training decreases computational complexity, memory usage, and communication cost by ∼4× and compute energy by a factor of ∼20×, while trading off less than a 1% accuracy for both IID and non-IID data. In particular, for higher skew values, we observe an increase in accuracy (by ∼0.5%) with low precision training, indicating the regularization effect of the quantization.

An Audit of Diagnostic Disparity between Intraoperative Frozen Section Diagnosis and Final Histopathological Diagnosis of Central Nervous System Lesions at a Tertiary Care Center

Introduction Evaluation of intraoperative squash smear and frozen section (FS) in central nervous system (CNS) neoplasms is consistently practiced for rapid assessment and has several advantages to its credence. It is an invaluable tool to ensure adequacy of tissue obtained to establish the diagnosis. Moreover, it aids in guiding the surgeon for critical decisions regarding the extent of resection. Although molecular markers have been integrated with morphology in the revised 2016 World Health Organization classification of brain tumors, precise morphological assessment still remains the foundation for the diagnosis and rapid intraoperative assessment of morphological details is equally critical and rewarding.

Objective This study aims to audit the diagnostic disparity between intraoperative diagnoses based on a combination of squash cytology and FS in cases of CNS lesions with gold standard, final diagnosis based on examination of formalin fixed paraffin embedded hematoxylin, and eosin-stained tissue sections.

Materials and Methods All intraoperative squash cytology and FS reported for CNS lesions from January 2017 to December 2020 were reviewed. The cases were categorized into three groups—group 1: when diagnosis of intraoperative diagnosis based on a combination of squash cytology and FS was same as the final histopathological diagnosis (concordant), group 2: partially concordant, and group 3: discordant cases.

Statistical Analysis Descriptive statistics was used to classify the data and diagnostic accuracy was calculated.

Results Complete concordance was present in 69.96% (191/273) cases, 20.1% (55/273) cases showed partial concordance, and 9.89% (27/273) cases were discordant with histopathological diagnosis. Out of the 27 discordant cases, misclassification of tumor type was the most common category (11 cases, 40%), followed by grading mismatch (7 cases, 25.9%), and misdiagnosis of tumor versus nontumor conditions (9 cases, 33.3%).

Conclusion Our study shows that combination of intraoperative squash cytology and FS shows a high percentage of accuracy in arriving at intraoperative diagnosis in cases of intracranial lesions. Regular audits of discordant cases should be conducted by surgeons and pathologists as part of a quality assurance measure to sensitize themselves with the potential pitfalls, minimizing misinterpretation and helping in providing a more conclusive opinion to the operating surgeons.

Clinical characteristics with inflammation profiling of long COVID and association with 1-year recovery following hospitalisation in the UK: a prospective observational study

BACKGROUND

No effective pharmacological or non-pharmacological interventions exist for patients with long COVID. We aimed to describe recovery 1 year after hospital discharge for COVID-19, identify factors associated with patient-perceived recovery, and identify potential therapeutic targets by describing the underlying inflammatory profiles of the previously described recovery clusters at 5 months after hospital discharge.

METHODS

The Post-hospitalisation COVID-19 study (PHOSP-COVID) is a prospective, longitudinal cohort study recruiting adults (aged ≥18 years) discharged from hospital with COVID-19 across the UK. Recovery was assessed using patient-reported outcome measures, physical performance, and organ function at 5 months and 1 year after hospital discharge, and stratified by both patient-perceived recovery and recovery cluster. Hierarchical logistic regression modelling was performed for patient-perceived recovery at 1 year. Cluster analysis was done using the clustering large applications k-medoids approach using clinical outcomes at 5 months. Inflammatory protein profiling was analysed from plasma at the 5-month visit. This study is registered on the ISRCTN Registry, ISRCTN10980107, and recruitment is ongoing.

FINDINGS

2320 participants discharged from hospital between March 7, 2020, and April 18, 2021, were assessed at 5 months after discharge and 807 (32·7%) participants completed both the 5-month and 1-year visits. 279 (35·6%) of these 807 patients were women and 505 (64·4%) were men, with a mean age of 58·7 (SD 12·5) years, and 224 (27·8%) had received invasive mechanical ventilation (WHO class 7-9). The proportion of patients reporting full recovery was unchanged between 5 months (501 [25·5%] of 1965) and 1 year (232 [28·9%] of 804). Factors associated with being less likely to report full recovery at 1 year were female sex (odds ratio 0·68 [95% CI 0·46-0·99]), obesity (0·50 [0·34-0·74]) and invasive mechanical ventilation (0·42 [0·23-0·76]). Cluster analysis (n=1636) corroborated the previously reported four clusters: very severe, severe, moderate with cognitive impairment, and mild, relating to the severity of physical health, mental health, and cognitive impairment at 5 months. We found increased inflammatory mediators of tissue damage and repair in both the very severe and the moderate with cognitive impairment clusters compared with the mild cluster, including IL-6 concentration, which was increased in both comparisons (n=626 participants). We found a substantial deficit in median EQ-5D-5L utility index from before COVID-19 (retrospective assessment; 0·88 [IQR 0·74-1·00]), at 5 months (0·74 [0·64-0·88]) to 1 year (0·75 [0·62-0·88]), with minimal improvements across all outcome measures at 1 year after discharge in the whole cohort and within each of the four clusters.

INTERPRETATION

The sequelae of a hospital admission with COVID-19 were substantial 1 year after discharge across a range of health domains, with the minority in our cohort feeling fully recovered. Patient-perceived health-related quality of life was reduced at 1 year compared with before hospital admission. Systematic inflammation and obesity are potential treatable traits that warrant further investigation in clinical trials.

FUNDING

UK Research and Innovation and National Institute for Health Research.

Transcervical Fibroid Ablation (TFA) with the Sonata System: Updated Review of a New Paradigm for Myoma Treatment

Purpose of Review

This paper provides an update of the recent evidence related to transcervical fibroid ablation (TFA) with the Sonata® System.

Recent Findings

An additional 27 papers representing over 400 women and more than 800 treated fibroids have been published. These demonstrate consistent, positive results, representing multiple prospective clinical trials, subgroup analyses, health economic analyses, case series, and systematic reviews of TFA. These include a 147-patient prospective clinical trial of TFA (the SONATA Clinical Trial) that demonstrated sustained symptom relief and an 8.2% cumulative reintervention rate through 3 years, a long-term study (VITALITY) confirming durable symptom relief with an 11.8% reintervention rate over > 5 years of mean follow-up, a clinical trial (OPEN) suggesting minimal potential for intrauterine adhesiogenesis post-TFA, preliminary results of a global registry (SAGE), and two subgroup analyses of TFA reporting favorable and safe outcomes in women with large fibroids > 5 cm in diameter. Three comparative health economic studies demonstrate favorable economic outcomes against both myomectomy and hysterectomy. A recent ACOG Practice Bulletin also noted equivalent outcomes for transcervical, laparoscopic, and transvaginal fibroid ablation.

Summary

Accumulated clinical evidence, including systematic reviews and longitudinal prospective clinical trials, continues to confirm the safety and efficacy of TFA in women with symptomatic fibroids, including myomata > 5 cm. As a transcervical treatment modality that can safely address all nonpedunculated uterine fibroid types, the continued evidence base confirms TFA as an innovative and useful treatment option that meets a significant unmet clinical need, including among underserved populations, delivering significant durable reductions in fibroid symptoms.

Application of cross-validation strategies to avoid overestimation of performance of 2D-QSAR models for the prediction of aquatic toxicity of chemical mixtures

The quantitative structure-activity relationship (QSAR) modelling of mixtures is not as simple as that for individual chemicals, and it needs additional care to avoid overestimation of the performance. In this research, we have developed a 2D-QSAR model using only 2D interpretable and reproducible descriptors to predict the aquatic toxicity of mixtures of polar and non-polar narcotic substances present in the environment. Partial least squares (PLS) regression has been used to model the response variable (log 1/EC50 against Photobacterium phosphoreum) and the structural features of 84 binary mixtures of polar and nonpolar narcotic toxicants complying with the Organization of Economic Co-operation and Development (OECD) protocols. The model was cross-validated by mixtures-out and compounds-out cross-validation to nullify the developmental bias. The reliability of prediction of the model has been judged by the Prediction Reliability Indicator (PRI) tool using a newly designed set. The new model is robust, reproducible, extremely predictive, easily interpretable, and can be used for reliable prediction of aquatic toxicity of any untested chemical mixtures within the applicability domain. We have additionally used a machine learning-based chemical read-across algorithm in this study to improve the quality of predictions for the toxicity of the mixtures with the modelled descriptors.

Noise resilient leaky integrate-and-fire neurons based on multi-domain spintronic devices

The capability of emulating neural functionalities efficiently in hardware is crucial for building neuromorphic computing systems. While various types of neuro-mimetic devices have been investigated, it remains challenging to provide a compact device that can emulate spiking neurons. In this work, we propose a non-volatile spin-based device for efficiently emulating a leaky integrate-and-fire neuron. By incorporating an exchange-coupled composite free layer in spin-orbit torque magnetic tunnel junctions, multi-domain magnetization switching dynamics is exploited to realize gradual accumulation of membrane potential for a leaky integrate-and-fire neuron with compact footprints. The proposed device offers significantly improved scalability compared with previously proposed spin-based neuro-mimetic implementations while exhibiting high energy efficiency and good controllability. Moreover, the proposed neuron device exhibits a varying leak constant and a varying membrane resistance that are both dependent on the magnitude of the membrane potential. Interestingly, we demonstrate that such device-inspired dynamic behaviors can be incorporated to construct more robust spiking neural network models, and find improved resiliency against various types of noise injection scenarios. The proposed spintronic neuro-mimetic devices may potentially open up exciting opportunities for the development of efficient and robust neuro-inspired computational hardware.

Identification of potential antivirals against 3CLpro enzyme for the treatment of SARS-CoV-2: A multi-step virtual screening study

The severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) outbreak is posing a serious public health threat worldwide in the form of COVD-19. Herein, we have performed two-dimensional quantitative structure-activity relationship (2D-QSAR) and three-dimensional pharmacophore modelling analysis employing inhibitors of 3-chymotrypsin-like protease (3CLpro), the leading protease that is crucial for the replication of SARS-CoV-2. The investigation aims to identify the important structural features responsible for the enzyme inhibition and the search for novel 3CLpro enzyme inhibitors as effective therapeutics for treating SARS-CoV-2. Furthermore, we carried out molecular docking studies using the most and least active compounds in the dataset, aiming to validate the contributions of various features as appeared in the QSAR models. Later, the stringently validated 2D-QSAR model was used to estimate the 3CLpro inhibitory activity of compounds from five chemical databases. Compounds with the significant predicted activity were then subjected to pharmacophore-based virtual screening to screen the top-rated compounds, which were then further subjected to molecular docking analysis, absorption, distribution, metabolism, excretion - toxicity (ADMET) profiling, and molecular dynamics (MD) simulation. The multi-step virtual screening analyses suggested that compounds CASAntiV-865453-58-3, CASAntiV-865453-40-3, and CASAntiV-2043031-84-9 could be used as effective therapeutic agents for the treatment of SARS-CoV-2.

QSAR modelling of organic dyes for their acute toxicity in Daphnia magna using 2D-descriptors

The present study reports quantitative structure-activity relationship (QSAR) models for 22 organic dyes spanning a broad chemical domain to predict their toxicity in Daphnia magna [log (1/EC₅₀)]. Only two-dimensional descriptors with clear physicochemical meaning were used to construct the QSAR models. The process of development, validation, and interpretation of models adheres to the stringent recommendations of the Organization for Economic Cooperation and Development (OECD) guidelines. In this study, the multi-layered stepwise regression method and linear discriminant analysis (LDA) method were employed for the deployment of regression - and classification-based models respectively; however, the final regression-based QSAR models were obtained through the partial least squares (PLS) regression. Additionally, the applicability domain of the developed models was verified. The constructed models should be applicable in the absence of toxicity data of new or untested dye structures, particularly when the compounds fall within the developed models' scope, and also implementable to develop more environmentally friendly alternatives.

BYANJON: A Ground Truth Preparation System for Online Handwritten Bangla Documents

The work reported in this article deals with the ground truth generation scheme for online handwritten Bangla documents at text-line, word, and stroke levels. The aim of the proposed scheme is twofold: firstly, to build a document level database so that future researchers can use the database to do research in this field. Secondly, the ground truth information will help other researchers to evaluate the performance of their algorithms developed for text-line extraction, word extraction, word segmentation, stroke recognition, and word recognition. The reported ground truth generation scheme starts with text-line extraction from the online handwritten Bangla documents, then words extraction from the text-lines, and finally segmentation of those words into basic strokes. After word segmentation, the basic strokes are assigned appropriate class labels by using modified distance-based feature extraction procedure and the MLP ( Multi-layer Perceptron ) classifier. The Unicode for the words are then generated from the sequence of stroke labels. XML files are used to store the stroke, word, and text-line levels ground truth information for the corresponding documents. The proposed system is semi-automatic and each step such as text-line extraction, word extraction, word segmentation, and stroke recognition has been implemented by using different algorithms. Thus, the proposed ground truth generation procedure minimizes huge manual intervention by reducing the number of mouse clicks required to extract text-lines, words from the document, and segment the words into basic strokes. The integrated stroke recognition module also helps to minimize the manual labor needed to assign appropriate stroke labels. The freely available and can be accessed at https://byanjon.herokuapp.com/ .

BlocTrain: Block-Wise Conditional Training and Inference for Efficient Spike-Based Deep Learning

Spiking neural networks (SNNs), with their inherent capability to learn sparse spike-based input representations over time, offer a promising solution for enabling the next generation of intelligent autonomous systems. Nevertheless, end-to-end training of deep SNNs is both compute- and memory-intensive because of the need to backpropagate error gradients through time. We propose BlocTrain, which is a scalable and complexity-aware incremental algorithm for memory-efficient training of deep SNNs. We divide a deep SNN into blocks, where each block consists of few convolutional layers followed by a classifier. We train the blocks sequentially using local errors from the classifier. Once a given block is trained, our algorithm dynamically figures out easy vs. hard classes using the class-wise accuracy, and trains the deeper block only on the hard class inputs. In addition, we also incorporate a hard class detector (HCD) per block that is used during inference to exit early for the easy class inputs and activate the deeper blocks only for the hard class inputs. We trained ResNet-9 SNN divided into three blocks, using BlocTrain, on CIFAR-10 and obtained 86.4% accuracy, which is achieved with up to 2.95× lower memory requirement during the course of training, and 1.89× compute efficiency per inference (due to early exit strategy) with 1.45× memory overhead (primarily due to classifier weights) compared to end-to-end network. We also trained ResNet-11, divided into four blocks, on CIFAR-100 and obtained 58.21% accuracy, which is one of the first reported accuracy for SNN trained entirely with spike-based backpropagation on CIFAR-100.

1664 A rare case of Stenotrophomas maltophilia affecting an ACL Endobutton: A Case Report and Review of the literature

Background

Stenotrophomas maltophilia (SM) is a ubiquitous opportunistic gram-negative bacillus known to cause hospital acquired infections, commonly in immunosuppressed individuals. However, occurrence of this microorganism is rarely reported in the musculoskeletal system. We report a case of a young healthy man with recurrent knee swelling caused by SM. This organism has been reported to form biofilms over orthopaedic implants, with increased resistance to antibiotic therapy, making it difficult to eradicate once established.

Case Report

A 34-year-old man with psoriasis and asthma presented with right knee swelling and pain, and no history of preceding trauma or fever. He has a past surgical history of anterior cruciate ligament tear and underwent reconstruction with an endobutton two years prior, and a lateral meniscal repair treated in the previous year. At presentation, he was apyrexial and was noted to have a moderate right knee effusion. Initial radiographic imaging identified a large joint effusion with no obvious evidence of bone destruction, laboratory studies also showed a raised white cell count of 11.4 *10^9/L, C-reactive protein of 54mg/L and aspiration cultures were negative. He subsequently underwent numerous washouts, and removal of endobutton with open synovial biopsy which showed SM. The organism was sensitive to cotrimoxazole. The patient had appropriate antibiotics with resolution of symptoms, however he still presented with recurrence.

Conclusions

The persistence of SM despite intervention makes it difficult to treat. Although a rare organism, surgeons should consider it as a differential in recurring knee effusion with orthopaedic implants, and institute early and appropriate management for best outcomes.

Quantifying the Brain Predictivity of Artificial Neural Networks With Nonlinear Response Mapping

Quantifying the similarity between artificial neural networks (ANNs) and their biological counterparts is an important step toward building more brain-like artificial intelligence systems. Recent efforts in this direction use neural predictivity, or the ability to predict the responses of a biological brain given the information in an ANN (such as its internal activations), when both are presented with the same stimulus. We propose a new approach to quantifying neural predictivity by explicitly mapping the activations of an ANN to brain responses with a non-linear function, and measuring the error between the predicted and actual brain responses. Further, we propose to use a neural network to approximate this mapping function by training it on a set of neural recordings. The proposed method was implemented within the TensorFlow framework and evaluated on a suite of 8 state-of-the-art image recognition ANNs. Our experiments suggest that the use of a non-linear mapping function leads to higher neural predictivity. Our findings also reaffirm the observation that the latest advances in classification performance of image recognition ANNs are not matched by improvements in their neural predictivity. Finally, we examine the impact of pruning, a widely used ANN optimization, on neural predictivity, and demonstrate that network sparsity leads to higher neural predictivity.

Using Embedded Feature Selection and CNN for Classification on CCD-INID-V1-A New IoT Dataset

As Internet of Things (IoT) networks expand globally with an annual increase of active devices, providing better safeguards to threats is becoming more prominent. An intrusion detection system (IDS) is the most viable solution that mitigates the threats of cyberattacks. Given the many constraints of the ever-changing network environment of IoT devices, an effective yet lightweight IDS is required to detect cyber anomalies and categorize various cyberattacks. Additionally, most publicly available datasets used for research do not reflect the recent network behaviors, nor are they made from IoT networks. To address these issues, in this paper, we have the following contributions: (1) we create a dataset from IoT networks, namely, the Center for Cyber Defense (CCD) IoT Network Intrusion Dataset V1 (CCD-INID-V1); (2) we propose a hybrid lightweight form of IDS—an embedded model (EM) for feature selection and a convolutional neural network (CNN) for attack detection and classification. The proposed method has two models: (a) RCNN: Random Forest (RF) is combined with CNN and (b) XCNN: eXtreme Gradient Boosting (XGBoost) is combined with CNN. RF and XGBoost are the embedded models to reduce less impactful features. (3) We attempt anomaly (binary) classifications and attack-based (multiclass) classifications on CCD-INID-V1 and two other IoT datasets, the detection_of_IoT_botnet_attacks_N_BaIoT dataset (Balot) and the CIRA-CIC-DoHBrw-2020 dataset (DoH20), to explore the effectiveness of these learning-based security models. Using RCNN, we achieved an Area under the Receiver Characteristic Operator (ROC) Curve (AUC) score of 0.956 with a runtime of 32.28 s on CCD-INID-V1, 0.999 with a runtime of 71.46 s on Balot, and 0.986 with a runtime of 35.45 s on DoH20. Using XCNN, we achieved an AUC score of 0.998 with a runtime of 51.38 s for CCD-INID-V1, 0.999 with a runtime of 72.12 s for Balot, and 0.999 with a runtime of 72.91 s for DoH20. Compared to KNN, XCNN required 86.98% less computational time, and RCNN required 91.74% less computational time to achieve equal or better accurate anomaly detections. We find XCNN and RCNN are consistently efficient and handle scalability well; in particular, 1000 times faster than KNN when dealing with a relatively larger dataset-Balot. Finally, we highlight RCNN and XCNN’s ability to accurately detect anomalies with a significant reduction in computational time. This advantage grants flexibility for the IDS placement strategy. Our IDS can be placed at a central server as well as resource-constrained edge devices. Our lightweight IDS requires low train time and hence decreases reaction time to zero-day attacks.

DeepRMethylSite: a deep learning based approach for prediction of arginine methylation sites in proteins

Methylation, which is one of the most prominent post-translational modifications on proteins, regulates many important cellular functions. Though several model-based methylation site predictors have been reported, all existing methods employ machine learning strategies, such as support vector machines and random forest, to predict sites of methylation based on a set of "hand-selected" features. As a consequence, the subsequent models may be biased toward one set of features. Moreover, due to the large number of features, model development can often be computationally expensive. In this paper, we propose an alternative approach based on deep learning to predict arginine methylation sites. Our model, which we termed DeepRMethylSite, is computationally less expensive than traditional feature-based methods while eliminating potential biases that can arise through features selection. Based on independent testing on our dataset, DeepRMethylSite achieved efficiency scores of 68%, 82% and 0.51 with respect to sensitivity (SN), specificity (SP) and Matthew's correlation coefficient (MCC), respectively. Importantly, in side-by-side comparisons with other state-of-the-art methylation site predictors, our method performs on par or better in all scoring metrics tested.

QSPR modelling for investigation of different properties of aminoglycoside-derived polymers using 2D descriptors

The quantitative structure-property relationship (QSPR) method is commonly used to predict different physicochemical characteristics of interest of chemical compounds with an objective to accelerate the process of design and development of novel chemical compounds in the biotechnology and healthcare industries. In the present report, we have employed a QSPR approach to predict the different properties of the aminoglycoside-derived polymers (i.e. polymer DNA binding and aminoglycoside-derived polymers mediated transgene expression). The final QSPR models were obtained using the partial least squares (PLS) regression approach using only specific categories of two-dimensional descriptors and subsequently evaluated considering different internationally accepted validation metrics. The proposed models are robust and non-random, demonstrating excellent predictive ability using test set compounds. We have also developed different kinds of consensus models using several validated individual models to improve the prediction quality for external set compounds. The present findings provide new insight for exploring the design of an aminoglycoside-derived polymer library based on different identified physicochemical properties as well as predict their property before their synthesis.

A deep learning based approach for prediction of Chlamydomonas reinhardtii phosphorylation sites

Protein phosphorylation, which is one of the most important post-translational modifications (PTMs), is involved in regulating myriad cellular processes. Herein, we present a novel deep learning based approach for organism-specific protein phosphorylation site prediction in Chlamydomonas reinhardtii, a model algal phototroph. An ensemble model combining convolutional neural networks and long short-term memory (LSTM) achieves the best performance in predicting phosphorylation sites in C. reinhardtii. Deemed Chlamy-EnPhosSite, the measured best AUC and MCC are 0.90 and 0.64 respectively for a combined dataset of serine (S) and threonine (T) in independent testing higher than those measures for other predictors. When applied to the entire C. reinhardtii proteome (totaling 1,809,304 S and T sites), Chlamy-EnPhosSite yielded 499,411 phosphorylated sites with a cut-off value of 0.5 and 237,949 phosphorylated sites with a cut-off value of 0.7. These predictions were compared to an experimental dataset of phosphosites identified by liquid chromatography-tandem mass spectrometry (LC–MS/MS) in a blinded study and approximately 89.69% of 2,663 C. reinhardtii S and T phosphorylation sites were successfully predicted by Chlamy-EnPhosSite at a probability cut-off of 0.5 and 76.83% of sites were successfully identified at a more stringent 0.7 cut-off. Interestingly, Chlamy-EnPhosSite also successfully predicted experimentally confirmed phosphorylation sites in a protein sequence (e.g., RPS6 S245) which did not appear in the training dataset, highlighting prediction accuracy and the power of leveraging predictions to identify biologically relevant PTM sites. These results demonstrate that our method represents a robust and complementary technique for high-throughput phosphorylation site prediction in C. reinhardtii. It has potential to serve as a useful tool to the community. Chlamy-EnPhosSite will contribute to the understanding of how protein phosphorylation influences various biological processes in this important model microalga.

Atlantoaxial Dislocation: Surgical Outcome following Anterior Retropharyngeal Approach to Odontoid with or without Fixation

Introduction Atlantoaxial dislocation (AAD) refers to a loss of stability between the atlas and axis vertebra. The conventional transoral approach and the anterior retropharyngeal approach (ARPA) are adequate for the upper cervical vertebral segments. This study was undertaken to evaluate the usefulness of the ARPA to the odontoid in patients of AAD.

Materials and Methods The study was undertaken on 20 patients admitted in Nil Ratan Sircar Medical College and Hospital, Kolkata, India, with functional disability secondary to AAD. These patients underwent surgery through ARPA to odontoid with or without fixation. Patients were analyzed between October 2014 and September 2016.

Results Maximum number of patients belonged to third decade of life. The male to female ratio was 1.5. The mean duration of symptoms was 10.86 months. Weakness of the upper and lower limbs predominated. About 65% patients had axial neck pain. Nine patients (45%) in total had difficulty in either bowel or bladder. Five patients presented with fracture odontoid and pannus formation of the odontoid process, while six had basilar invagination. One patient underwent anterior odontoid screw fixation and the other 19 patients underwent anterior retropharyngeal odontoidectomy with posterior fixation. Two patients expired in the present study. Most of the patients had improvement in Nurick grade during follow-up. Five patients had transient throat pain and dysphagia. Three patients had superficial surgical site infection. One patient had postoperative cerebrospinal fluid leak.

Conclusion The ARPA to odontoid is a feasible approach for decompression and fixation of the odontoid in AAD cases.

Multi-Level Neuromorphic Devices Built on Emerging Ferroic Materials: A Review

Achieving multi-level devices is crucial to efficiently emulate key bio-plausible functionalities such as synaptic plasticity and neuronal activity, and has become an important aspect of neuromorphic hardware development. In this review article, we focus on various ferromagnetic (FM) and ferroelectric (FE) devices capable of representing multiple states, and discuss the usage of such multi-level devices for implementing neuromorphic functionalities. We will elaborate that the analog-like resistive states in ferromagnetic or ferroelectric thin films are due to the non-coherent multi-domain switching dynamics, which is fundamentally different from most memristive materials involving electroforming processes or significant ion motion. Both device fundamentals related to the mechanism of introducing multilevel states and exemplary implementations of neural functionalities built on various device structures are highlighted. In light of the non-destructive nature and the relatively simple physical process of multi-domain switching, we envision that ferroic-based multi-state devices provide an alternative pathway toward energy efficient implementation of neuro-inspired computing hardware with potential advantages of high endurance and controllability.

Implicit adversarial data augmentation and robustness with Noise-based Learning

We introduce a Noise-based Learning (NoL) approach for training neural networks that are intrinsically robust to adversarial attacks. We find that the learning of random noise introduced with the input with the same loss function used during posterior maximization, improves a model's adversarial resistance. We show that the learnt noise performs implicit adversarial data augmentation boosting a model's adversary generalization capability. We evaluate our approach's efficacy and provide a simplistic visualization tool for understanding adversarial data, using Principal Component Analysis. We conduct comprehensive experiments on prevailing benchmarks such as MNIST, CIFAR10, CIFAR100, Tiny ImageNet and show that our approach performs remarkably well against a wide range of attacks. Furthermore, combining NoL with state-of-the-art defense mechanisms, such as adversarial training, consistently outperforms prior techniques in both white-box and black-box attacks.

Shallow Convolutional Neural Network for COVID-19 Outbreak Screening Using Chest X-rays

Among radiological imaging data, Chest X-rays (CXRs) are of great use in observing COVID-19 manifestations. For mass screening, using CXRs, a computationally efficient AI-driven tool is the must to detect COVID-19-positive cases from non-COVID ones. For this purpose, we proposed a light-weight Convolutional Neural Network (CNN)-tailored shallow architecture that can automatically detect COVID-19-positive cases using CXRs, with no false negatives. The shallow CNN-tailored architecture was designed with fewer parameters as compared to other deep learning models. The shallow CNN-tailored architecture was validated using 321 COVID-19-positive CXRs. In addition to COVID-19-positive cases, another set of non-COVID-19 5856 cases (publicly available, source: Kaggle) was taken into account, consisting of normal, viral, and bacterial pneumonia cases. In our experimental tests, to avoid possible bias, 5-fold cross-validation was followed, and both balanced and imbalanced datasets were used. The proposed model achieved the highest possible accuracy of 99.69%, sensitivity of 1.0, where AUC was 0.9995. Furthermore, the reported false positive rate was only 0.0015 for 5856 COVID-19-negative cases. Our results stated that the proposed CNN could possibly be used for mass screening. Using the exact same set of CXR collection, the current results were better than other deep learning models and major state-of-the-art works.

TRAF6-IRF5 kinetics, TRIF, and biophysical factors drive synergistic innate responses to particle-mediated MPLA-CpG co-presentation

Innate immune responses to pathogens are driven by co-presentation of multiple pathogen-associated molecular patterns (PAMPs). Combinations of PAMPs can trigger synergistic immune responses, but the underlying molecular mechanisms of synergy are poorly understood. Here, we used synthetic particulate carriers co-loaded with monophosphoryl lipid A (MPLA) and CpG as pathogen-like particles (PLPs) to dissect the signaling pathways responsible for dual adjuvant immune responses. PLP-based co-delivery of MPLA and CpG to GM-CSF-driven mouse bone marrow-derived antigen-presenting cells (BM-APCs) elicited synergistic interferon-β (IFN-β) and interleukin-12p70 (IL-12p70) responses, which were strongly influenced by the biophysical properties of PLPs. Mechanistically, we found that MyD88 and interferon regulatory factor 5 (IRF5) were necessary for IFN-β and IL-12p70 production, while TRIF signaling was required for the synergistic response. Both the kinetics and magnitude of downstream TRAF6 and IRF5 signaling drove the synergy. These results identify the key mechanisms of synergistic Toll-like receptor 4 (TLR4)-TLR9 co-signaling in mouse BM-APCs and underscore the critical role of signaling kinetics and biophysical properties on the integrated response to combination adjuvants.

Pre-Operative MRI in evaluation of Perianal fistulas

Background: Perianal fistula is a common and often an extremely distressing disorder. MRI is a preferred modality for the preoperative assessment of perianal fistulas. Aims and Objective: To evaluate the accuracy and predictive values of pre-operative MRI in diagnosing severity of perianal fistulas and the conditions associated with it like internal anal opening, secondary tract, abscess and supralevator extension. Materials and Methods: In this study, 44 patients with clinically suspected perianal fistulas underwent MRI for the evaluation of severity and presence of the associated conditions. St James’s University Hospital Classification was used to grade the perianal fistula. Pre-operative MRI grading was compared with the surgical findings in the 26 operated patients. Results: Out of 44 clinically suspected perianal fistulas, 41 (93%) were diagnosed as perianal fistulas by MRI. The most common type was grade 1 seen in 12(27.3%) patients followed by grade 4 fistula seen in 10(22.7%) patients. All 26 patients evaluated by surgical findings for severity of perianal fistulas were found to be graded in identical grades by pre-operative MRI. MRI was 100% sensitive in picking up all five grades. Conclusion: MRI helps in the accurate delineation of fistulous tract, identification of secondary tract and abscess.

Roadmap of spin-orbit torques

Spin-orbit torque (SOT) is an emerging technology that enables the efficient manipulation of spintronic devices. The initial processes of interest in SOTs involved electric fields, spin-orbit coupling, conduction electron spins and magnetization. More recently interest has grown to include a variety of other processes that include phonons, magnons, or heat. Over the past decade, many materials have been explored to achieve a larger SOT efficiency. Recently, holistic design to maximize the performance of SOT devices has extended material research from a nonmagnetic layer to a magnetic layer. The rapid development of SOT has spurred a variety of SOT-based applications. In this Roadmap paper, we first review the theories of SOTs by introducing the various mechanisms thought to generate or control SOTs, such as the spin Hall effect, the Rashba-Edelstein effect, the orbital Hall effect, thermal gradients, magnons, and strain effects. Then, we discuss the materials that enable these effects, including metals, metallic alloys, topological insulators, two-dimensional materials, and complex oxides. We also discuss the important roles in SOT devices of different types of magnetic layers, such as magnetic insulators, antiferromagnets, and ferrimagnets. Afterward, we discuss device applications utilizing SOTs. We discuss and compare three-terminal and two-terminal SOT-magnetoresistive random-access memories (MRAMs); we mention various schemes to eliminate the need for an external field. We provide technological application considerations for SOT-MRAM and give perspectives on SOT-based neuromorphic devices and circuits. In addition to SOT-MRAM, we present SOT-based spintronic terahertz generators, nano-oscillators, and domain wall and skyrmion racetrack memories. This paper aims to achieve a comprehensive review of SOT theory, materials, and applications, guiding future SOT development in both the academic and industrial sectors.

Roadmap on emerging hardware and technology for machine learning

Recent progress in artificial intelligence is largely attributed to the rapid development of machine learning, especially in the algorithm and neural network models. However, it is the performance of the hardware, in particular the energy efficiency of a computing system that sets the fundamental limit of the capability of machine learning. Data-centric computing requires a revolution in hardware systems, since traditional digital computers based on transistors and the von Neumann architecture were not purposely designed for neuromorphic computing. A hardware platform based on emerging devices and new architecture is the hope for future computing with dramatically improved throughput and energy efficiency. Building such a system, nevertheless, faces a number of challenges, ranging from materials selection, device optimization, circuit fabrication and system integration, to name a few. The aim of this Roadmap is to present a snapshot of emerging hardware technologies that are potentially beneficial for machine learning, providing the Nanotechnology readers with a perspective of challenges and opportunities in this burgeoning field.