Abstract: The behavior of metals across a broad frequency range from microwave to ultraviolet frequencies is of interest in plasmonics, nanophotonics, and metamaterials. Depending on the frequency, losses of collective excitations in metals can be predominantly classical resistive effects or Landau damping. In this context, we present first-principles calculations that capture all of the significant microscopic mechanisms underlying surface plasmon decay and predict the initial excited carrier distributions so generated. Specifically, we include ab initio predictions of phonon-assisted optical excitations in metals, which are critical to bridging the frequency range between resistive losses at low frequencies and direct interband transitions at high frequencies. In the commonly used plasmonic materials, gold, silver, copper, and aluminum, we find that resistive losses compete with phonon-assisted carrier generation below the interband threshold, but hot carrier generation via direct transitions dominates above thre...

Abstract: Cell-free Massive MIMO is considered as a promising technology for satisfying the increasing number of users and high rate expectations in beyond-5G networks. The key idea is to let many distributed access points (APs) communicate with all users in the network, possibly by using joint coherent signal processing. The aim of this paper is to provide the first comprehensive analysis of this technology under different degrees of cooperation among the APs. Particularly, the uplink spectral efficiencies of four different cell-free implementations are analyzed, with spatially correlated fading and arbitrary linear processing. It turns out that it is possible to outperform conventional Cellular Massive MIMO and small cell networks by a wide margin, but only using global or local minimum mean-square error (MMSE) combining. This is in sharp contrast to the existing literature, which advocates for maximum-ratio combining. Also, we show that a centralized implementation with optimal MMSE processing not only maximizes the SE but largely reduces the fronthaul signaling compared to the standard distributed approach. This makes it the preferred way to operate Cell-free Massive MIMO networks. Non-linear decoding is also investigated and shown to bring negligible improvements.

Abstract: CT imaging is crucial for diagnosis, assessment and staging COVID-19 infection. Follow-up scans every 3-5 days are often recommended for disease progression. It has been reported that bilateral and peripheral ground glass opacification (GGO) with or without consolidation are predominant CT findings in COVID-19 patients. However, due to lack of computerized quantification tools, only qualitative impression and rough description of infected areas are currently used in radiological reports. In this paper, a deep learning (DL)-based segmentation system is developed to automatically quantify infection regions of interest (ROIs) and their volumetric ratios w.r.t. the lung. The performance of the system was evaluated by comparing the automatically segmented infection regions with the manually-delineated ones on 300 chest CT scans of 300 COVID-19 patients. For fast manual delineation of training samples and possible manual intervention of automatic results, a human-in-the-loop (HITL) strategy has been adopted to assist radiologists for infection region segmentation, which dramatically reduced the total segmentation time to 4 minutes after 3 iterations of model updating. The average Dice simiarility coefficient showed 91.6% agreement between automatic and manual infaction segmentations, and the mean estimation error of percentage of infection (POI) was 0.3% for the whole lung. Finally, possible applications, including but not limited to analysis of follow-up CT scans and infection distributions in the lobes and segments correlated with clinical findings, were discussed.

Abstract: We introduce a framework that abstracts Reinforcement Learning (RL) as a sequence modeling problem. This allows us to draw upon the simplicity and scalability of the Transformer architecture, and associated advances in language modeling such as GPT-x and BERT. In particular, we present Decision Transformer, an architecture that casts the problem of RL as conditional sequence modeling. Unlike prior approaches to RL that fit value functions or compute policy gradients, Decision Transformer simply outputs the optimal actions by leveraging a causally masked Transformer. By conditioning an autoregressive model on the desired return (reward), past states, and actions, our Decision Transformer model can generate future actions that achieve the desired return. Despite its simplicity, Decision Transformer matches or exceeds the performance of state-of-the-art model-free offline RL baselines on Atari, OpenAI Gym, and Key-to-Door tasks.

Abstract: Lightweight conductive porous graphene/thermoplastic polyurethane (TPU) foams with ultrahigh compressibility were successfully fabricated by using the thermal induced phase separation (TISP) technique. The density and porosity of the foams were calculated to be about 0.11 g cm−3 and 90% owing to the porous structure. Compared with pure TPU foams, the addition of graphene could effectively increase the thickness of the cell wall and hinder the formation of small holes, leading to a robust porous structure with excellent compression property. Meanwhile, the cell walls with small holes and a dendritic structure were observed due to the flexibility of graphene, endowing the foam with special positive piezoresistive behaviors and peculiar response patterns with a deflection point during the cyclic compression. This could effectively enhance the identifiability of external compression strain when used as piezoresistive sensors. In addition, larger compression sensitivity was achieved at a higher compression rate. Due to high porosity and good elasticity of TPU, the conductive foams demonstrated good compressibility and stable piezoresistive sensing signals at a strain of up to 90%. During the cyclic piezoresistive sensing test under different compression strains, the conductive foam exhibited good recoverability and reproducibility after the stabilization of cyclic loading. All these suggest that the fabricated conductive foam possesses great potential to be used as lightweight, flexible, highly sensitive, and stable piezoresistive sensors.

Abstract: How well does a classic deep net architecture like AlexNet or VGG19 classify on a standard dataset such as CIFAR-10 when its “width”— namely, number of channels in convolutional layers, and number of nodes in fully-connected internal layers — is allowed to increase to infinity? Such questions have come to the forefront in the quest to theoretically understand deep learning and its mysteries about optimization and generalization. They also connect deep learning to notions such as Gaussian processes and kernels. A recent paper [Jacot et al., 2018] introduced the Neural Tangent Kernel (NTK) which captures the behavior of fully-connected deep nets in the infinite width limit trained by gradient descent; this object was implicit in some other recent papers. An attraction of such ideas is that a pure kernel-based method is used to capture the power of a fully-trained deep net of infinite width. The current paper gives the first efficient exact algorithm for computing the extension of NTK to convolutional neural nets, which we call Convolutional NTK (CNTK), as well as an efficient GPU implementation of this algorithm. This results in a significant new benchmark for performance of a pure kernel-based method on CIFAR-10, being 10% higher than the methods reported in [Novak et al., 2019], and only 6% lower than the performance of the corresponding finite deep net architecture (once batch normalization etc. are turned off). Theoretically, we also give the first non-asymptotic proof showing that a fully-trained sufficiently wide net is indeed equivalent to the kernel regression predictor using NTK.

Abstract: Tyrosinase is a multi-copper enzyme which is widely distributed in different organisms and plays an important role in the melanogenesis and enzymatic browning. Therefore, its inhibitors can be attractive in cosmetics and medicinal industries as depigmentation agents and also in food and agriculture industries as antibrowning compounds. For this purpose, many natural, semi-synthetic and synthetic inhibitors have been developed by different screening methods to date. This review has focused on the tyrosinase inhibitors discovered from all sources and biochemically characterised in the last four decades.

Abstract: Just as decentralization communication systems lead to the creation of the Internet, today a new technology — the blockchain — has the potential to decentralize the way we store data and manage information, potentially leading to a reduced role for one of the most important regulatory actors in our society: the middleman. Blockchain technology enables the creation of decentralized currencies, self-executing digital contracts (smart contracts) and intelligent assets that can be controlled over the Internet (smart property). The blockchain also enables the development of new governance systems with more democratic or participatory decision-making, and decentralized (autonomous) organizations that can operate over a network of computers without any human intervention. These applications have lead many to compare the blockchain to the Internet, with accompanying predictions that this technology will shift the balance of power away from centralized authorities in the field of communications, business, and even politics or law.In this Article, we explore the benefits and drawbacks of this emerging decentralized technology and argue that its widespread deployment will lead to expansion of a new subset of law, which we term Lex Cryptographia: rules administered through self-executing smart contracts and decentralized (autonomous) organizations. As blockchain technology becomes widely adopted, centralized authorities, such as governmental agencies and large multinational corporations, could lose the ability to control and shape the activities of disparate people through existing means. As a result, there will be an increasing need to focus on how to regulate blockchain technology and how to shape the creation and deployment of these emerging decentralized organizations in ways that have yet to be explored under current legal theory.

Abstract: A twentyfold improvement in the accuracy of a single ytterbium ion atomic clock is achieved using the ion's electric octupole transition.

Abstract: Poly(ethylene terephthalate) (PET) is one of the most abundantly produced synthetic polymers and is accumulating in the environment at a staggering rate as discarded packaging and textiles. The properties that make PET so useful also endow it with an alarming resistance to biodegradation, likely lasting centuries in the environment. Our collective reliance on PET and other plastics means that this buildup will continue unless solutions are found. Recently, a newly discovered bacterium, Ideonella sakaiensis 201-F6, was shown to exhibit the rare ability to grow on PET as a major carbon and energy source. Central to its PET biodegradation capability is a secreted PETase (PET-digesting enzyme). Here, we present a 0.92 A resolution X-ray crystal structure of PETase, which reveals features common to both cutinases and lipases. PETase retains the ancestral α/β-hydrolase fold but exhibits a more open active-site cleft than homologous cutinases. By narrowing the binding cleft via mutation of two active-site residues to conserved amino acids in cutinases, we surprisingly observe improved PET degradation, suggesting that PETase is not fully optimized for crystalline PET degradation, despite presumably evolving in a PET-rich environment. Additionally, we show that PETase degrades another semiaromatic polyester, polyethylene-2,5-furandicarboxylate (PEF), which is an emerging, bioderived PET replacement with improved barrier properties. In contrast, PETase does not degrade aliphatic polyesters, suggesting that it is generally an aromatic polyesterase. These findings suggest that additional protein engineering to increase PETase performance is realistic and highlight the need for further developments of structure/activity relationships for biodegradation of synthetic polyesters.

Abstract: Importance More than 80% of patients with heart failure with preserved ejection fraction (HFPEF), the most common form of heart failure among older persons, are overweight or obese. Exercise intolerance is the primary symptom of chronic HFPEF and a major determinant of reduced quality of life (QOL). Objective To determine whether caloric restriction (diet) or aerobic exercise training (exercise) improves exercise capacity and QOL in obese older patients with HFPEF. Design, Setting, and Participants Randomized, attention-controlled, 2 × 2 factorial trial conducted from February 2009 through November 2014 in an urban academic medical center. Of 577 initially screened participants, 100 older obese participants (mean [SD]: age, 67 years [5]; body mass index, 39.3 [5.6]) with chronic, stable HFPEF were enrolled (366 excluded by inclusion and exclusion criteria, 31 for other reasons, and 80 declined participation). Interventions Twenty weeks of diet, exercise, or both; attention control consisted of telephone calls every 2 weeks. Main Outcomes and Measures Exercise capacity measured as peak oxygen consumption (Vo 2 , mL/kg/min; co–primary outcome) and QOL measured by the Minnesota Living with Heart Failure (MLHF) Questionnaire (score range: 0–105, higher scores indicate worse heart failure–related QOL; co–primary outcome). Results Of the 100 enrolled participants, 26 participants were randomized to exercise; 24 to diet; 25 to exercise + diet; 25 to control. Of these, 92 participants completed the trial. Exercise attendance was 84% (SD, 14%) and diet adherence was 99% (SD, 1%). By main effects analysis, peak Vo 2 was increased significantly by both interventions: exercise, 1.2 mL/kg body mass/min (95% CI, 0.7 to 1.7), P P 2 (joint effect, 2.5 mL/kg/min). There was no statistically significant change in MLHF total score with exercise and with diet (main effect: exercise, −1 unit [95% CI, −8 to 5], P = .70; diet, −6 units [95% CI, −12 to 1], P = .08). The change in peak Vo 2 was positively correlated with the change in percent lean body mass ( r = 0.32; P = .003) and the change in thigh muscle:intermuscular fat ratio ( r = 0.27; P = .02). There were no study-related serious adverse events. Body weight decreased by 7% (7 kg [SD, 1]) in the diet group, 3% (4 kg [SD, 1]) in the exercise group, 10% (11 kg [SD, 1] in the exercise + diet group, and 1% (1 kg [SD, 1]) in the control group. Conclusions and Relevance Among obese older patients with clinically stable HFPEF, caloric restriction or aerobic exercise training increased peak Vo 2 , and the effects may be additive. Neither intervention had a significant effect on quality of life as measured by the MLHF Questionnaire. Trial Registration clinicaltrials.gov Identifier:NCT00959660

Abstract: Semantic image segmentation is an essential component of modern autonomous driving systems, as an accurate understanding of the surrounding scene is crucial to navigation and action planning. Current state-of-the-art approaches in semantic image segmentation rely on pre-trained networks that were initially developed for classifying images as a whole. While these networks exhibit outstanding recognition performance (i.e., what is visible?), they lack localization accuracy (i.e., where precisely is something located?). Therefore, additional processing steps have to be performed in order to obtain pixel-accurate segmentation masks at the full image resolution. To alleviate this problem we propose a novel ResNet-like architecture that exhibits strong localization and recognition performance. We combine multi-scale context with pixel-level accuracy by using two processing streams within our network: One stream carries information at the full image resolution, enabling precise adherence to segment boundaries. The other stream undergoes a sequence of pooling operations to obtain robust features for recognition. The two streams are coupled at the full image resolution using residuals. Without additional processing steps and without pre-training, our approach achieves an intersection-over-union score of 71.8% on the Cityscapes dataset.

Abstract: To the Editor: A cluster of unexplained pneumonia cases was reported by the Peopleʼs Republic of China to the World Health Organization (WHO) on 31 December, 2019. The etiology for this outbreak was a novel coronavirus named severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which was responsible for the Corona Virus Disease 2019 (COVID-19). Singapore confirmed its first imported case on 23 January 2020 and local transmission was detected on 4 February, 2020. As of 28 February, 2020, Singapore had 96 confirmed cases of COVID-19 infection. SARS-CoV-2 was confirmed by real time reverse transcriptase-polymerase chain reaction (RT-PCR), performed on respiratory samples of these patients. A majority of 69 out of these 96 patients were treated at the National Centre for Infectious Diseases (NCID). We herein present a detailed analysis of the hematological parameters of the COVID-19 patients at the NCID (see Table 1). Of the 69 patients that had been admitted to the NCID, 26 patients were still hospitalized, and 43 patients had been discharged as of 28 February 2020. Also, 67 patients had at least one complete blood count (CBC) performed during inpatient stay; 65 patients had CBC performed on day of admission. We analyzed the hematological indices of all COVID-19 infected patients from day 1 of admission until 28 February 2020. We obtained data from the Laboratory Information System (LIS) exclusively which provided information on the age, gender, ethnicity and location of each patient. We divided the patients into two groups; ICU and non-ICU patients. Additionally, flow cytometry on lymphocyte subsets was performed from 24 to 28 February 2020, on a subgroup of nine COVID-19 patients; five ICU patients and four non-ICU patients (with six normal individual blood samples as controls). Immunophenotyping was performed using a Becton Dickinson FACSCanto II Flow analyzer. Most patients were of Chinese ethnicity (89.5%), while the minority were ofMalay (4.5%), Indian (1.5%) and other ethnicities (4.5%). Just 9 out of the 67 (13.4%) patients required ICU care. Notably, ICU patients were about a decade older than the non-ICU patients; the median age of ICU patients was 54 years old while the median age of non-ICU patients was 42 years old (P = .02). On admission, leukopenia was observed in 19 patients (29.2%) with only one patient presenting with severe leukopenia (WBC < 2 × 10/L). Lymphopenia featured in 24 patients (36.9%) with 19 having moderate lymphopenia (absolute lymphocyte count [ALC] 0.5-1 × 10/L), and five with severe lymphopenia (ALC < 0.5 × 10/L). Most patients had normal platelet counts, with 13 patients (20.0%) having mild thrombocytopenia (platelet count 100-150 × 10/L). Peripheral blood film review showed that a higher number of patients (69%) who were lymphopenic had the presence of a few reactive lymphocytes, of which a subset appeared lymphoplasmacytoid. This contrasts with the severe acute respiratory syndrome (SARS) outbreak in 2003 where reactive lymphocytes were not observed in a study on Haematologic parameters in SARS in Singapore and only in 15.2% of cases in a similar Hong Kong study. Our analysis revealed that on admission, most patients had a normal CBC (normal Hb, WBC and platelet count) and lactate dehydrogenase (LDH). And, no patient presented with moderate or severe thrombocytopenia that is frequently observed in other viral illnesses such as dengue fever which is endemic in our region. However, 28% of all patients presented with lymphopenia (ALC < 1 × 10/L). This number is significantly smaller compared to 63% of patients in Wuhan, China, and 42% of patients outside of Wuhan who presented with lymphopenia. This disparity in numbers may in part be reflective of the extent of epidemiological data availablewithin the surveillance pyramid in those regions. Those requiring ICU care had a lower ALC and higher LDH. These were findings also reported by Huang et al on the characteristics of COVID-19 patients inWuhan, China. Lymphopenia has beenwell described in retrospective analysis of patients in Hong Kong and Singapore afflicted with SARS-CoV in 2003, and was associated with adverse outcomes and ICU stay. Lymphopenia featured prominently in our COVID-19 ICU groupwith amedian nadir ALC of 0.4× 10/L, compared to 1.2 × 10/L in the non-ICU group. Monitoring of such hematologic parameters may help to identify patients whomay need ICU care. An ALC approaching severe lymphopenia of <0.6 × 10/L may possibly be considered as one of the indicators for early admission for supportive care in the ICU. Between the ICU (n = 9) and non-ICU (n = 58) patients, using Fisherʼs exact tests, we found that admission ALC and LDH stood out as discriminating laboratory indices with a P value of <.001 and .005 respectively. The ICU patients in general presented with more profound lymphopenia with seven out of nine being lymphopenic; four of whom had severe lymphopenia. Note, LDH was performed for 4 out of the 9 ICU patients on admission, and all four cases had a raised LDH with a median value of 1684 U/L (reference range 270-550 U/L). Comparatively non-ICU patients tend to present with a normal LDH, median value 401 U/L; with only five out of 26 non-ICU patients presenting with a Received: 2 March 2020 Revised: 3 March 2020 Accepted: 3 March 2020

Abstract: Background Septic shock is characterized by dysregulation of the host response to infection, with circulatory, cellular, and metabolic abnormalities. We hypothesized that therapy with hydrocortisone plus fludrocortisone or with drotrecogin alfa (activated), which can modulate the host response, would improve the clinical outcomes of patients with septic shock. Methods In this multicenter, double-blind, randomized trial with a 2-by-2 factorial design, we evaluated the effect of hydrocortisone-plus-fludrocortisone therapy, drotrecogin alfa (activated), the combination of the three drugs, or their respective placebos. The primary outcome was 90-day all-cause mortality. Secondary outcomes included mortality at intensive care unit (ICU) discharge and hospital discharge and at day 28 and day 180 and the number of days alive and free of vasopressors, mechanical ventilation, or organ failure. After drotrecogin alfa (activated) was withdrawn from the market, the trial continued with a two-group parallel d...

Abstract: Viability is the ability of a supply chain (SC) to maintain itself and survive in a changing environment through a redesign of structures and replanning of performance with long-term impacts. In this paper, we theorize a new notion-the viable supply chain (VSC). In our approach, viability is considered as an underlying SC property spanning three perspectives, i.e., agility, resilience, and sustainability. The principal ideas of the VSC model are adaptable structural SC designs for supply-demand allocations and, most importantly, establishment and control of adaptive mechanisms for transitions between the structural designs. Further, we demonstrate how the VSC components can be categorized across organizational, informational, process-functional, technological, and financial structures. Moreover, our study offers a VSC framework within an SC ecosystem. We discuss the relations between resilience and viability. Through the lens and guidance of dynamic systems theory, we illustrate the VSC model at the technical level. The VSC model can be of value for decision-makers to design SCs that can react adaptively to both positive changes (i.e., the agility angle) and be able to absorb negative disturbances, recover and survive during short-term disruptions and long-term, global shocks with societal and economical transformations (i.e., the resilience and sustainability angles). The VSC model can help firms in guiding their decisions on recovery and re-building of their SCs after global, long-term crises such as the COVID-19 pandemic. We emphasize that resilience is the central perspective in the VSC guaranteeing viability of the SCs of the future. Emerging directions in VSC research are discussed.

Abstract: An accumulation of unfolded or misfolded proteins in the endoplasmic reticulum (ER) leads to stress conditions. To mitigate such circumstances, stressed cells activate a homeostatic intracellular signaling network cumulatively called the unfolded protein response (UPR), which orchestrates the recuperation of ER function. Macroautophagy (hereafter autophagy), an intracellular lysosome-mediated bulk degradation pathway for recycling and eliminating wornout proteins, protein aggregates, and damaged organelles, has also emerged as an essential protective mechanism during ER stress. These 2 systems are dynamically interconnected, and recent investigations have revealed that ER stress can either stimulate or inhibit autophagy. However, the stress-associated molecular cues that control the changeover switch between induction and inhibition of autophagy are largely obscure. This review summarizes the crosstalk between ER stress and autophagy and their signaling networks mainly in mammalian-based systems. Additionally, we highlight current knowledge on selective autophagy and its connection to ER stress.

Abstract: Low dimensional materials including quantum dots, nanowires, 2D materials, and so forth have attracted increasing research interests for electronic and optoelectronic devices in recent years. Photogating, which is usually observed in photodetectors based on low dimensional materials and their hybrid structures, is demonstrated to play an important role. Photogating is considered as a way of conductance modulation through photoinduced gate voltage instead of simply and totally attributing it to trap states. This review first focuses on the gain of photogating and reveals the distinction from conventional photoconductive effect. The trap- and hybrid-induced photogating including their origins, formations, and characteristics are subsequently discussed. Then, the recent progress on trap- and hybrid-induced photogating in low dimensional photodetectors is elaborated. Though a high gain bandwidth product as high as 109 Hz is reported in several cases, a trade-off between gain and bandwidth has to be made for this type of photogating. The general photogating is put forward according to another three reported studies very recently. General photogating may enable simultaneous high gain and high bandwidth, paving the way to explore novel high-performance photodetectors.

Abstract: We present a novel method for scalable and precise certification of deep neural networks. The key technical insight behind our approach is a new abstract domain which combines floating point polyhedra with intervals and is equipped with abstract transformers specifically tailored to the setting of neural networks. Concretely, we introduce new transformers for affine transforms, the rectified linear unit (ReLU), sigmoid, tanh, and maxpool functions. We implemented our method in a system called DeepPoly and evaluated it extensively on a range of datasets, neural architectures (including defended networks), and specifications. Our experimental results indicate that DeepPoly is more precise than prior work while scaling to large networks. We also show how to combine DeepPoly with a form of abstraction refinement based on trace partitioning. This enables us to prove, for the first time, the robustness of the network when the input image is subjected to complex perturbations such as rotations that employ linear interpolation.

Abstract: Neural-based multi-task learning has been successfully used in many real-world large-scale applications such as recommendation systems. For example, in movie recommendations, beyond providing users movies which they tend to purchase and watch, the system might also optimize for users liking the movies afterwards. With multi-task learning, we aim to build a single model that learns these multiple goals and tasks simultaneously. However, the prediction quality of commonly used multi-task models is often sensitive to the relationships between tasks. It is therefore important to study the modeling tradeoffs between task-specific objectives and inter-task relationships. In this work, we propose a novel multi-task learning approach, Multi-gate Mixture-of-Experts (MMoE), which explicitly learns to model task relationships from data. We adapt the Mixture-of-Experts (MoE) structure to multi-task learning by sharing the expert submodels across all tasks, while also having a gating network trained to optimize each task. To validate our approach on data with different levels of task relatedness, we first apply it to a synthetic dataset where we control the task relatedness. We show that the proposed approach performs better than baseline methods when the tasks are less related. We also show that the MMoE structure results in an additional trainability benefit, depending on different levels of randomness in the training data and model initialization. Furthermore, we demonstrate the performance improvements by MMoE on real tasks including a binary classification benchmark, and a large-scale content recommendation system at Google.

Abstract: Voice interfaces are becoming more ubiquitous and are now the primary input method for many devices. We explore in this paper how they can be attacked with hidden voice commands that are unintelligible to human listeners but which are interpreted as commands by devices. We evaluate these attacks under two different threat models. In the black-box model, an attacker uses the speech recognition system as an opaque oracle. We show that the adversary can produce difficult to understand commands that are effective against existing systems in the black-box model. Under the white-box model, the attacker has full knowledge of the internals of the speech recognition system and uses it to create attack commands that we demonstrate through user testing are not understandable by humans. We then evaluate several defenses, including notifying the user when a voice command is accepted; a verbal challenge-response protocol; and a machine learning approach that can detect our attacks with 99.8% accuracy.

Abstract: Thermal performance curves (TPCs), which quantify how an ectotherm's body temperature (Tb ) affects its performance or fitness, are often used in an attempt to predict organismal responses to climate change. Here, we examine the key - but often biologically unreasonable - assumptions underlying this approach; for example, that physiology and thermal regimes are invariant over ontogeny, space and time, and also that TPCs are independent of previously experienced Tb. We show how a critical consideration of these assumptions can lead to biologically useful hypotheses and experimental designs. For example, rather than assuming that TPCs are fixed during ontogeny, one can measure TPCs for each major life stage and incorporate these into stage-specific ecological models to reveal the life stage most likely to be vulnerable to climate change. Our overall goal is to explicitly examine the assumptions underlying the integration of TPCs with Tb , to develop a framework within which empiricists can place their work within these limitations, and to facilitate the application of thermal physiology to understanding the biological implications of climate change.

Abstract: Population growth combined with increasingly limited resources of arable land and fresh water has resulted in a need for alternative protein sources. Macroalgae (seaweed) and microalgae are examples of under-exploited “crops”. Algae do not compete with traditional food crops for space and resources. This review details the characteristics of commonly consumed algae, as well as their potential for use as a protein source based on their protein quality, amino acid composition, and digestibility. Protein extraction methods applied to algae to date, including enzymatic hydrolysis, physical processes, and chemical extraction and novel methods such as ultrasound-assisted extraction, pulsed electric field, and microwave-assisted extraction are discussed. Moreover, existing protein enrichment methods used in the dairy industry and the potential of these methods to generate high value ingredients from algae, such as bioactive peptides and functional ingredients are discussed. Applications of algae in human nutrition, animal feed, and aquaculture are examined.

Abstract: We investigate a principle way to progressively mine discriminative object regions using classification networks to address the weakly-supervised semantic segmentation problems. Classification networks are only responsive to small and sparse discriminative regions from the object of interest, which deviates from the requirement of the segmentation task that needs to localize dense, interior and integral regions for pixel-wise inference. To mitigate this gap, we propose a new adversarial erasing approach for localizing and expanding object regions progressively. Starting with a single small object region, our proposed approach drives the classification network to sequentially discover new and complement object regions by erasing the current mined regions in an adversarial manner. These localized regions eventually constitute a dense and complete object region for learning semantic segmentation. To further enhance the quality of the discovered regions by adversarial erasing, an online prohibitive segmentation learning approach is developed to collaborate with adversarial erasing by providing auxiliary segmentation supervision modulated by the more reliable classification scores. Despite its apparent simplicity, the proposed approach achieves 55.0% and 55.7% mean Intersection-over-Union (mIoU) scores on PASCAL VOC 2012 val and test sets, which are the new state-of-the-arts.

Abstract: In late December 2019 an outbreak of a novel coronavirus (SARS-CoV-2) causing severe pneumonia (COVID-19) was reported in Wuhan, Hubei Province, China. A common finding in most COVID-19 patients is high D-dimer levels which are associated with a worse prognosis. We aimed to evaluate coagulation abnormalities via traditional tests and whole blood thromboelastometry profiles in a group of 22 (mean age 67 ± 8 years, M:F 20:2) consecutive patients admitted to the Intensive Care Unit of Padova University Hospital for acute respiratory failure due to COVID-19. Cases showed significantly higher fibrinogen and D-dimer plasma levels versus healthy controls (p < 0.0001 in both comparisons). Interestingly enough, markedly hypercoagulable thromboelastometry profiles were observed in COVID-19 patients, as reflected by shorter Clot Formation Time (CFT) in INTEM (p = 0.0002) and EXTEM (p = 0.01) and higher Maximum Clot Firmness (MCF) in INTEM, EXTEM and FIBTEM (p < 0.001 in all comparisons). In conclusion, COVID-19 patients with acute respiratory failure present a severe hypercoagulability rather than consumptive coagulopathy. Fibrin formation and polymerization may predispose to thrombosis and correlate with a worse outcome.