Search

Abstract While the activities of certain proteases promote proteostasis and prevent neurodegeneration-associated phenotypes, the protease cathepsin B (CTSB) enhances proteotoxicity in Alzheimer’s disease (AD) model mice, and its levels are elevated in brains of AD patients. How CTSB exacerbates the toxicity of the AD-causing Amyloid β (Aβ) peptide is controversial. Using an activity-based probe, aging-altering interventions and the nematode C. elegans, we discovered that the CTSB CPR-6 promotes Aβ proteotoxicity but mitigates the toxicity of polyQ stretches. While the knockdown of cpr-6 does not affect lifespan, it alleviates Aβ toxicity by reducing the expression of swsn-3 and elevating the level of the protein SMK-1, both involved in the regulation of aging. These observations unveil a mechanism by which CTSB aggravates Aβ–mediated toxicity, indicate that it plays opposing roles in the face of distinct proteotoxic insults and highlight the importance of tailoring specific remedies for distinct neurodegenerative disorders.

High-energy gamma-ray radiography has exceptional penetration ability and has become an indispensable nondestructive testing (NDT) tool in various fields. For high-energy photons, point projection radiography is almost the only feasible imaging method, and its spatial resolution is primarily constrained by the size of the gamma-ray source. In conventional industrial applications, gamma-ray sources are commonly based on electron beams driven by accelerators, utilizing the process of bremsstrahlung radiation. The size of the gamma-ray source is dependent on the dimensional characteristics of the electron beam. Extensive research has been conducted on various advanced accelerator technologies that have the potential to greatly improve spatial resolution in NDT. In our investigation of laser-driven gamma-ray sources, a spatial resolution of about 90 µm is achieved when the areal density of the penetrated object is 120 g/cm2. A virtual source approach is proposed to optimize the size of the gamma-ray source used for imaging, with the aim of maximizing spatial resolution. In this virtual source approach, the gamma ray can be considered as being emitted from a virtual source within the convertor, where the equivalent gamma-ray source size in imaging is much smaller than the actual emission area. On the basis of Monte Carlo simulations, we derive a set of evaluation formulas for virtual source scale and gamma-ray emission angle. Under optimal conditions, the virtual source size can be as small as 15 µm, which can significantly improve the spatial resolution of high-penetration imaging to less than 50 µm.

Background: Insufficient clinicians' auscultation ability delays the diagnosis and treatment of valvular heart disease (VHD); artificial intelligence provides a solution to compensate for the insufficiency in auscultation ability by distinguishing between heart murmurs and normal heart sounds. However, whether artificial intelligence can automatically diagnose VHD remains unknown. Our objective was to use deep learning to process and compare raw heart sound data to identify patients with VHD requiring intervention. Methods: Heart sounds from patients with VHD and healthy controls were collected using an electronic stethoscope. Echocardiographic findings were used as the gold standard for this study. According to the chronological order of enrollment, the early-enrolled samples were used to train the deep learning model, and the late-enrollment samples were used to validate the results. Results: The final study population comprised 499 patients (354 in the algorithm training group and 145 in the result validation group). The sensitivity, specificity, and accuracy of the deep-learning model for identifying various VHDs ranged from 71.4 to 100.0%, 83.5–100.0%, and 84.1–100.0%, respectively; the best diagnostic performance was observed for mitral stenosis, with a sensitivity of 100.0% (31.0–100.0%), a specificity of 100% (96.7–100.0%), and an accuracy of 100% (97.5–100.0%). Conclusions: Based on raw heart sound data, the deep learning model effectively identifies patients with various types of VHD who require intervention and assists in the screening, diagnosis, and follow-up of VHD.

BackgroundEmpirical evidence has shown that light therapy (LT) can reduce depression symptoms by stimulating circadian rhythms. However, there is skepticism and inconclusive results, along with confusion regarding dosing. The purpose of this study is to quantify light as a stimulus for the circadian system and create a dose-response relationship that can help reduce maladies among adolescents and young adults (AYAs). This will provide a reference for light exposure and neural response, which are crucial in the neuropsychological mechanism of light intervention. The study also aims to provide guidance for clinical application.MethodsThe latest quantitative model of CLA (circadian light) and CSt,f (circadian stimulus) was adopted to quantify light dose for circadian phototransduction in youth depression-related light therapy. Articles published up to 2023 through Web of Science, Cochrane Library, Medline (OVID), CINAHL, APA PsycINFO, Embase, and Scholars were retrieved. A meta-analysis of 31 articles (1,031 subjects) was performed using Stata17.0, CMA3.0 (comprehensive meta-analysis version 3.0) software, and Python 3.9 platform for light therapy efficacy comparison and dose-response quantification.ResultsUnder various circadian stimulus conditions (0.1 < CSt,f < 0.7) of light therapy (LT), malady reductions among AYAs were observed (pooled SMD = −1.59, 95%CI = −1.86 to −1.32; z = −11.654, p = 0.000; I2 = 92.8%), with temporal pattern (p = 0.044) and co-medication (p = 0.000) suggested as main heterogeneity sources. For the efficacy advantage of LT with a higher circadian stimulus that is assumed to be influenced by visualization, co-medication, disease severity, and time pattern, sets of meta-analysis among random-controlled trials (RCTs) found evidence for significant efficacy of circadian-active bright light therapy (BLT) over circadian-inactive dim red light (SMD = −0.65, 95% CI = −0.96 to −0.34; z = −4.101, p = 0.000; I2 = 84.9%) or circadian-active dimmer white light (SMD = −0.37, 95% CI = −0.68 to −0.06; z = −2.318, p = 0.02; I2 = 33.8%), whereas green-blue, circadian-active BLT showed no significant superiority over circadian-inactive red/amber light controls (SMD = −0.21, 95% CI = −0.45 to 0.04; z = −2.318, p = 0.099; I2 = 0%). Overall, circadian-active BLT showed a greater likelihood of clinical response than dim light controls, with increased superiority observed with co-medication. For pre-to-post-treatment amelioration and corresponding dose-response relationship, cumulative duration was found more influential than other categorical (co-medication, severity, study design) or continuous (CSt,f) variables. Dose-response fitting indicated that the therapeutic effect would reach saturation among co-medicated patients at 32–42 days (900–1,000 min) and 58–59 days (1,100–1,500 min) among non-medicated AYAs. When exerting high circadian stimulus of light therapy (0.6 < CSt,f < 0.7), there was a significantly greater effect size in 1,000–1,500 min of accumulative duration than 1,500 min of duration, indicating a threshold for practical guidance.LimitationsThe results have been based on limited samples and influenced by a small sample effect. The placebo effect could not be ignored.ConclusionsAlthough the superiority of LT with higher circadian stimulus over dimmer light controls remains unproven, greater response potentials of circadian-active BLT have been noticed among AYAs, taking co-medication, disease severity, time pattern, and visual characteristics into consideration. The dose-response relationship with quantified circadian stimulus and temporal pattern had been elaborated under various conditions to support clinical depression treatment and LT device application in the post-pandemic era.

Anabolic androgenic steroids (AASs) are usually illegally added to animal feed because they can significantly promote animal growth and increase carcasses’ leanness, which threatens the safety of animal-derived foods and indirectly hazards human health. This study aimed to establish an ultra-high performance liquid chromatography-tandem mass spectrometry (UHPLC–MS/MS) method for the simultaneous detection of twelve AAS residues in livestock and poultry meat. The homogenized samples were extracted with acetonitrile containing 1% acetic acid (v/v) and purified using the one-step extraction column. After concentration using nitrogen, the residues were redissolved in acetonitrile and then quantified with an external standard method using UHPLC–MS/MS. The results showed that the above-mentioned method had a satisfactory linear correlation (R2 ≥ 0.9903) with a concentration range of 1–100 μg/L, and the limits of detection (LODs) and quantification (LOQs) were 0.03–0.33 μg/kg and 0.09–0.90 μg/kg, respectively. With the intraday and interday precision less than 15%, the average recoveries of pork, beef, lamb, and chicken, at different spiked levels, ranged from 68.3 to 93.3%, 68.0 to 99.4%, 71.6 to 109.8%, and 70.5 to 97.7%, respectively. Overall, the established method is validated, precise, and capable of the high-throughput determination of the residues of twelve AASs in livestock and poultry meat.

X-ray source based on the Thomson scattering of laser interacting with energetic electron beams features high photon energy, small spot size, and good collimation. However, the photon number is insufficient for practical application because of the small cross section of the Thomson scattering. To solve this problem, here, we replace a traditional Gaussian laser pulse with a flying focus laser pulse to extend interaction length and restrain nonlinear effects. Simulation results show that the scattered photon number can be increased by about 25 and 2 times for high and low energy lasers, respectively. In particular, a 1010 photon number can be generated with a 10 J flying focus laser pulse, and the energy spread can also be greatly reduced for high energy laser, from a broad spectrum to a monoenergetic peak. Combining these two advantages, the peak spectrum brightness of x ray is 3 × 108 photons/keV at 240 keV, which is about three orders of magnitude more than the traditional case.

In recent years, the role of video games in enhancing brain plasticity and learning ability has been verified, and this learning transfer is known as the “learning to learn” effect of video game training. At the same time, against the background of healthy lighting, the influence of non-visual effects of light environment on the human rhythmic system has been gradually confirmed. As a special operation form of Visual Display Terminal (VDT) operation, video game training has a high dependence on VDT equipment and the VDT screen, and the background usually has a huge difference in brightness. Compared with the light environment of ordinary operation space, the light environment of VDT operation space is more complex. This complex light environment's non-visual effects cause human emotions, alertness, fatigue, cognitive ability, and other changes, which may affect the efficiency of the “learning to learn” effect of video game training. This article focuses on the impact of the light environment in the VDT workspace on the “learning to learn” effect of video game training. It first traces the factors that trigger the “learning to learn” effect of video game training, that is, the improvement of people's attention, perception, and cognitive ability. Then, the influencing mechanism and the evaluation method of the VDT workspace space light environment on the human rhythm system are discussed based on the basic theory of photobiological effect. In addition, the VDT display lighting light time pattern, photophysical properties, regulation, and protection mechanism on the human rhythm system are studied to demonstrate the VDT workspace light environment's special characteristics. Finally, combined with the progress of artificial lighting technology and the research results of health lighting, given the “learning to learn” effect of video game training, some thoughts on the design of the light environment of the workplace and future research directions are presented.

Laser wakefield acceleration, a highly nonlinear process, can provide high-quality electron beams and betatron radiation. However, it has always been a challenge to optimize the photon number of betatron radiation because X-ray production is the secondary process followed by electron generation and many influential parameters are highly coupled with each other. To balance these parameters and optimize target value, we use a machine-learning algorithm called Bayesian optimization that ignores the complicated middle process and so is suitable to solve this kind of problems. Without designed initialization, we optimize betatron radiation in experiments by changing plasma density and laser focal position. The X-ray photon number doubles compared with initial condition within 10 iterations, which shows the rapidity of this algorithm. PIC simulation also illustrates the complexity of the optimization and reveals the reason why X-ray photon number increases.

This paper attempts to realize the balance between humans and ecology in designing the nighttime light environment of urban parks by clarifying the influence of nighttime artificial light on the ecosystem of urban parks. Firstly, we reviewed the effects of nighttime artificial light on individual predation and reproduction of animals and personal growth and reproduction of plants. Secondly, we discuss the impact of individual changes caused by artificial lighting on ecosystem function at the ecosystem and analyze its advantages and disadvantages. The results showed that nighttime artificial light had a double-sided impact on the ecosystem, which would hurt the ecosystem function, but had a positive effect on the green space, which lacked natural light and had high plant density. This paper focuses on the areas with increased application of artificial lighting and rich species of animals and plants in night cities, such as urban forest parks and urban green spaces. It discusses how to reduce the intrusion of artificial lighting on ecosystems and how to make better use of the positive effect of artificial light.

Internal mildewed nutmeg is difficult to perceive without cutting the nutmeg open and examining it carefully, which poses a significant risk to public health. At present, macroscopic identification and chromatographic analysis are applied to determine whether nutmeg is moldy or not. However, the former relies on a human panel, with the disadvantages of subjectivity and empirical dependence, whilst the latter is generally time-consuming and requires organic solvents. Therefore, it is urgent to develop a rapid and feasible approach for evaluating the quality and predicting mildew in nutmeg. In this study, the quality and odor characteristics of five groups of nutmeg samples with different degrees of mildew were analyzed by using the responses of an electronic nose combined with chemical profiling. The main physicochemical indicators, such as the levels of α-pinene, β-pinene, elemicin, and dehydro-di-isoeugenol, were determined. The results revealed that the contents of α-pinene, β-pinene, and elemicin changed significantly with the extension of storage time. Through the use of an electronic nose and HS–GC–MS technology to assess the overall odor characteristics of nutmeg samples, it was found that the production of volatile organic compounds (VOCs) such as ammonia/organic amines, carbon monoxide, ethanol, and hydrogen sulfide, as well as changes in the terpene and phenylpropene components of the nutmeg itself, may be the material basis for the changes in odor. The accuracy of the qualitative classification model for the degree of mildew in nutmeg was higher than 90% according to the electronic nose data combined with different machine learning algorithms. Quantitative models were established for predicting the contents of the chemical components, and models based on a BP neural network (BPNN), the support vector machine (SVM), and the random forest algorithm (RF) all showed good performance in predicting the concentrations of these chemical components, except for dehydro-di-isoeugenol. The BPNN performed effectively in predicting the storage time of nutmeg on the basis of the E-nose’s responses, with an RMSE and R2 of 0.268 and 0.996 for the training set, and 0.317 and 0.993 for the testing set, respectively. The results demonstrated that the responses of the electronic nose (E-nose) had a high correlation with the internal quality of nutmeg. This work proposes a quick and non-destructive evaluation method for the quality of nutmeg, which has high accuracy in discriminating between different degrees of mold in nutmeg and is conducive to early detection and warning of moldy phenomena.

How to balance the artistry and comfort of the light environment in an architecturalized immersive new media public art installation (AINMPAI) to ensure the safety of the audiences’ visual safety has become a new issue. The lack of corresponding lighting design standards makes this issue more challenging. This paper explores the appropriate luminance range of the irregularly curved LED screens of the AINMPAI with high-luminance natural light as the background. The influence of the audience behavior mode on the brightness threshold of the LED screens in the AINMPAI under the background of Internet communication is discussed. Through software simulation and field measurement, the effectiveness of the design measures based on the local characteristics of the work is verified. The overall average luminance, the regional luminance, the partial luminance, and the corresponding luminance contrast of the inner screen surface were measured at every 10° change in the solar altitude angle during daytime. The nighttime light environment parameters and the temperature of the device throughout the day were also measured. A total of 487 visitors were interviewed for subjective evaluation of the comfort of the light and thermal environments. The results show that: (1) the overall average luminance cannot wholly describe the actual luminance of the critical parts of the special-shaped curved screen in the device and that two indicators, the regional luminance and the partial average luminance, need to be added; (2) the maximum brightness limit of LED screens in the daytime can be 1000 cd/m2 but at night it should be controlled within 200 cd/m2; (3) natural light is the main factor that causes the high average brightness and low contrast of the daytime device screen; (4) the recommended indicators for such artistic installations should balance the absolute values of average brightness, artistic effect, and transmission priority. This study can provide foundational data and a methodological reference for establishing AINMPAI light environment design guidelines or recommended standards based on audiences’ visual safety.

Compared with the gold standard, polysomnography (PSG), and silver standard, actigraphy, contactless consumer sleep-tracking devices (CCSTDs) are more advantageous for implementing large-sample and long-period experiments in the field and out of the laboratory due to their low price, convenience, and unobtrusiveness. This review aimed to examine the effectiveness of CCSTDs application in human experiments. A systematic review and meta-analysis (PRISMA) of their performance in monitoring sleep parameters were conducted (PROSPERO: CRD42022342378). PubMed, EMBASE, Cochrane CENTRALE, and Web of Science were searched, and 26 articles were qualified for systematic review, of which 22 provided quantitative data for meta-analysis. The findings show that CCSTDs had a better accuracy in the experimental group of healthy participants who wore mattress-based devices with piezoelectric sensors. CCSTDs’ performance in distinguishing waking from sleeping epochs is as good as that of actigraphy. Moreover, CCSTDs provide data on sleep stages that are not available when actigraphy is used. Therefore, CCSTDs could be an effective alternative tool to PSG and actigraphy in human experiments.

In recent years, the domestic and international trade volumes of Chinese medicinal materials (CMMs) keep increasing. By the end of 2019, the total amount of exported CMMs reached as high as US $1.137 billion, while imported was US $2.155 billion. A stable and controllable quality system of CMMs apparently becomes the most important issue, which needs multifaceted collaboration from harvesting CMMs at a proper season to storing CMMs at a proper temperature. However, due to imperfect storage conditions, different kinds of deteriorations are prone to occur, for instance, get moldy or rancid, which not only causes a huge waste of CMM resources but also poses a great threat to clinical medication safety and public health. The key issue is to quickly and accurately distinguish deteriorated CMM samples so as to avoid consuming low-quality or even harmful CMMs. However, some attention has been paid to study the changing quality of deteriorated CMMs and a suitable method for identifying them. In this study, as a medicine and food material which easily becomes rancid, armeniacae semen amarum (ASA) was chosen as a research objective, and experimental ASA samples of different rancidness degrees were collected. Then, various kinds of analytical methods and technologies were applied to explore the changing rules of ASA quality and figure out the key indicators for the quality evaluation of ASA in the rancid process, including the human panel, colorimeter, electronic nose, and GC/MS. This study aims to analyze the correlation between the external morphological features and the inner chemical compounds, to find out the specific components from “quantitative change” to “qualitative change” in the process of “getting rancid,” and to discover the dynamic changes in the aforementioned key indicators at different stages of rancidness. The results showed since ASA samples began to get rancid with the extension of storage time, morphological features, namely, surface color and smell, changed significantly, and the degree of rancidness further deepened at the same time. Based on macroscopic identification accomplished via the human panel, ASA samples with varying degrees of rancidness were divided into four groups. The result of colorimeter analysis was in agreement with that of the human panel, as well as the determination of the amygdalin content and peroxide value. Moreover, there were obvious differences in the amygdalin content and peroxide value among ASA samples with different rancidness degrees. With a higher degree of rancidness, the content of amygdalin decreased, while the peroxide value increased significantly. The rancidness degree of ASA has a negative correlation with the amygdalin content and a positive correlation with the peroxide value. The newly discovered nonanal and 2-bromopropiophenone in rancid ASA samples may be the key components of “rancidity smell,” and these two components would be the exclusive components that trigger “quantitative change” to “qualitative change” in the process of rancidness of ASA. This study sheds light on studying the internal mechanism of “rancidness” of CMMs and provides an important basis for the effective storage and safe medication of easy-to-get rancid herbs, and it also plays an important foundation for the establishment of a stable and controllable quality system for CMMs.

Abstract Background Modification of RNAs, particularly at the terminals, is critical for various essential cell processes; for example, uridylation is implicated in tumorigenesis, proliferation, stem cell maintenance, and immune defense against viruses and retrotransposons. Ribosomal RNAs can be regulated by antisense ribosomal siRNAs (risiRNAs), which downregulate pre-rRNAs through the nuclear RNAi pathway in Caenorhabditis elegans. However, the biogenesis and regulation of risiRNAs remain obscure. Previously, we showed that 26S rRNAs are uridylated at the 3′-ends by an unknown terminal polyuridylation polymerase before the rRNAs are degraded by a 3′ to 5′ exoribonuclease SUSI-1(ceDIS3L2). Results Here, we found that CDE-1, one of the three C.elegans polyuridylation polymerases (PUPs), is specifically involved in suppressing risiRNA production. CDE-1 localizes to perinuclear granules in the germline and uridylates Argonaute-associated 22G-RNAs, 26S, and 5.8S rRNAs at the 3′-ends. Immunoprecipitation followed by mass spectrometry (IP-MS) revealed that CDE-1 interacts with SUSI-1(ceDIS3L2). Consistent with these results, both CDE-1 and SUSI-1(ceDIS3L2) are required for the inheritance of RNAi. Conclusions This work identified a rRNA surveillance machinery of rRNAs that couples terminal polyuridylation and degradation.

Background: Magnolia bark (Magnolia Officinalis REHD. & WILS. and Magnolia officinalis REHD. & WILS. VAR. biloba REHD. & WILS, Hou Po in Chinese), is widely applied in clinical prescriptions and Chinese patent medicines. Origin place is a crucial factor affecting the quality of Hou Po, and chemical composition is an important index for evaluating its quality, which is closely related to its clinical efficacy. This study aims to develop a novel method for rapidly, accurately and comprehensively identifying the origin places of Hou Po and predicting the contents of its important chemical components. Methods: High performance liquid chromatography was used to analyze the contents of magnolol and honokiol and ultra-performance liquid chromatography the contents of magnocurarine and magnoflorine. The cold soak method was used to determine the contents of water-soluble extracts. The E-nose and colorimeter were used to determine the odor and color characteristics, respectively, of the collected Hou Po samples. Results: Using several statistical algorithms, different discriminant models based on the E-nose and colorimeter data were established to distinguish the origin place of Hou-Po and predict the chemical components of honokiol, magnolol, magnocurarine, magnoflorine and water-soluble extracts. The results showed that the Random Forest classifier combined with the ten-fold cross-validation method provided the highest classification accuracy for origin place, accounting for 99.53% among these models. The correlation coefficients between predicted and experimental values of the five chemical components were all higher than 0.96. Conclusion: This study has indicated that the electronic nose and colorimeter are promising methods for evaluating the quality of Chinese herbal medicines both qualitatively and quantitatively.

Recently end-to-end (E2E) automatic speech recognition (ASR) models have achieved promising performance. However, existing models tend to adopt increasing model sizes and suffer from expensive resource consumption for real-world applications. To compress E2E ASR models and obtain smaller model sizes, we propose a comprehensive compression platform named LWMD (light-weight model designing), which consists of two essential parts: a light-weight architecture search (LWAS) framework and a differentiable structured pruning (DSP) algorithm. On the one hand, the LWAS framework adopts the neural architecture search (NAS) technique to automatically search light-weight architectures for E2E ASR models. By integrating different architecture topologies of existing models together, LWAS designs a topology-fused search space. Furthermore, combined with the E2E ASR training criterion, LWAS develops a resource-aware search algorithm to select light-weight architectures from the search space. On the other hand, given the searched architectures, the DSP algorithm performs structured pruning to reduce parameter numbers further. With a Gumbel re-parameter trick, DSP builds a stronger correlation between the pruning criterion and the model performance than conventional pruning methods. And an attention-similarity loss function is further developed for better performance. On two mandarin datasets, Aishell-1 and HKUST, the compression results are well evaluated and analyzed to demonstrate the effectiveness of the LWMD platform.

β-agonists are illegally added to animal feed because they can greatly increase carcasses’ leanness, which impairs the safety of animal-derived foods and indirectly endangers human health. This study aimed to develop an ultrahigh-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) method for determining sixteen β-agonists in livestock meat. The homogenized samples were subjected to enzymatic hydrolysis using β-glucuronidase/sulfatase at 40 °C for 2 h, extracted with acetonitrile containing 1% acetic acid (v/v), and purified by the one-step Qvet-AG extraction column. The residue was redissolved by 0.1% aqueous formic acid/methanol (9:1, v/v) after blow-drying by nitrogen, and then determined by UHPLC-MS/MS. The results demonstrated that the well linearity was in the range of 0.1–50 μg/L with the correlation coefficient (R2) ≥0.9928, and the limits of detection (LOD) and quantification (LOQ) were 0.01–0.11 μg/kg and 0.04–0.38 μg/kg, respectively. With intraday and interday relative standard deviations (RSDs) being less than 10%, the average recoveries of pork, beef, and lamb at various spiked levels ranged from 62.62–115.93%, 61.35–106.34%, and 62.00–111.83%, respectively. In conclusion, the established method is simple, efficient, sensitive, and suitable for the simultaneous detection of several β-agonist residues in livestock meat.

Single shot gamma-induced positron annihilation spectroscopy (GiPS) based on a laser wakefield accelerator is proposed. In this spectroscopy, a large number of gamma rays generated by laser-wakefield accelerated electrons are injected into a sample in a very short time (

BackgroundThe global outbreak of coronavirus disease 2019 (COVID-19) has turned into a worldwide public health crisis and caused more than 100,000,000 severe cases. Progressive lymphopenia, especially in T cells, was a prominent clinical feature of severe COVID-19. Activated HLA-DR+CD38+ CD8+ T cells were enriched over a prolonged period from the lymphopenia patients who died from Ebola and influenza infection and in severe patients infected with SARS-CoV-2. However, the CD38+HLA-DR+ CD8+ T population was reported to play contradictory roles in SARS-CoV-2 infection.MethodsA total of 42 COVID-19 patients, including 32 mild or moderate and 10 severe or critical cases, who received care at Beijing Ditan Hospital were recruited into this retrospective study. Blood samples were first collected within 3 days of the hospital admission and once every 3–7 days during hospitalization. The longitudinal flow cytometric data were examined during hospitalization. Moreover, we evaluated serum levels of 45 cytokines/chemokines/growth factors and 14 soluble checkpoints using Luminex multiplex assay longitudinally.ResultsWe revealed that the HLA-DR+CD38+ CD8+ T population was heterogeneous, and could be divided into two subsets with distinct characteristics: HLA-DR+CD38dim and HLA-DR+CD38hi. We observed a persistent accumulation of HLA-DR+CD38hi CD8+ T cells in severe COVID-19 patients. These HLA-DR+CD38hi CD8+ T cells were in a state of overactivation and consequent dysregulation manifested by expression of multiple inhibitory and stimulatory checkpoints, higher apoptotic sensitivity, impaired killing potential, and more exhausted transcriptional regulation compared to HLA-DR+CD38dim CD8+ T cells. Moreover, the clinical and laboratory data supported that only HLA-DR+CD38hi CD8+ T cells were associated with systemic inflammation, tissue injury, and immune disorders of severe COVID-19 patients.ConclusionsOur findings indicated that HLA-DR+CD38hi CD8+ T cells were correlated with disease severity of COVID-19 rather than HLA-DR+CD38dim population.

BackgroundAmomi fructus is a famous traditional Chinese medicine (TCM) that can exert beneficial effects during the treatment of gastrointestinal diseases and is used widely in China and other countries in Southeast Asia. However, the nonvolatile active ingredients that are present in the water extractions from A. fructus used to treat gastrointestinal diseases have yet to be elucidated. The goal of this study was to identify the nonvolatile active ingredients of A. fructus.MethodsWe used an in situ single-pass intestinal perfusion (SPIP) model to identify the active ingredients of A. fructus that play significant roles in gastrointestinal absorption. In addition, we developed a high-performance liquid chromatography (HPLC) method to identify key fractions in intestinal outflow perfusate.ResultsNineteen components were identified in a water extraction from A. fructus; these exhibited different absorption capabilities in different intestinal segments. Of these, six components were determined by the newly developed HPLC method: catechin, vanillic acid, epicatechin, polydatin, isoquercitrin, and quercitrin.ConclusionsThe current study aimed to identify the active ingredients present in water extractions prepared from A. fructus in a single-intestinal perfusate from rats. Our findings provide an experimental basis to explain the pharmacodynamic actions of A. fructus.

Network embedding task aims at learning low-dimension latent representations of vertices while preserving the structure of a network simultaneously. Most existing network embedding methods mainly focus on static networks, which extract and condense the network information without temporal information. However, in the real world, networks keep evolving, where the linkage states between the same vertex pairs at consequential timestamps have very close correlations. In this paper, we propose to study the network embedding problem and focus on modeling the linkage evolution in the dynamic network setting. To address this problem, we propose a deep dynamic network embedding method. More specifically, the method utilizes the historical information obtained from the network snapshots at past timestamps to learn latent representations of the future network. In the proposed embedding method, the objective function is carefully designed to incorporate both the network internal and network dynamic transition structures. Extensive empirical experiments prove the effectiveness of the proposed model on various categories of real-world networks, including a human contact network, a bibliographic network, and e-mail networks. Furthermore, the experimental results also demonstrate the significant advantages of the method compared with both the state-of-the-art embedding techniques and several existing baseline methods.

Link prediction in dynamic networks aims to predict edges according to historical linkage status. It is inherently difficult because of the linear/non-linear transformation of underlying structures. The problem of efficiently performing dynamic link inference is extremely challenging due to the scale of networks and different evolving patterns. Most previous approaches for link prediction are based on members' similarity and supervised learning methods. However, research work on investigating hidden patterns of dynamic social networks is rarely conducted. In this paper, we propose a novel framework that incorporates a deep learning method, i.e., temporal restricted Boltzmann machine, and a machine learning approach, i.e., gradient boosting decision tree. The proposed model is capable of modeling each link's evolving patterns. We also propose a novel transformation for input matrix, which significantly reduces the computational complexity and makes our algorithm scalable to large networks. Extensive experiments demonstrate that the proposed method outperforms the existing state-of-the-art algorithms on real-world dynamic networks.

Deep neural networks (DNNs) have been shown to be effective for single sound source localization in shallow water environments. However, multiple source localization is a more challenging task because of the interactions among multiple acoustic signals. This paper proposes a framework for multiple source localization on underwater horizontal arrays using deep neural networks. The two-stage DNNs are adopted to determine both the directions and ranges of multiple sources successively. A feed-forward neural network is trained for direction finding, while the long short term memory recurrent neural network is used for source ranging. Particularly, in the source ranging stage, we perform subarray beamforming to extract features of sources that are detected by the direction finding stage, because subarray beamforming can enhance the mixed signal to the desired direction while preserving the horizontal-longitudinal correlations of the acoustic field. In this way, a universal model trained in the single-source scenario can be applied to multi-source scenarios with arbitrary numbers of sources. Both simulations and experiments in a range-independent shallow water environment of SWellEx-96 Event S5 are given to demonstrate the effectiveness of the proposed method.

As demonstrated in hybrid connectionist temporal classification (CTC)/Attention architecture, joint training with a CTC objective is very effective to solve the misalignment problem existing in the attention-based end-to-end automatic speech recognition (ASR) framework. However, the CTC output relies only on the current input, which leads to the hard alignment issue. To address this problem, this paper proposes the time-restricted attention CTC/Attention architecture, which integrates an attention mechanism with the CTC branch. “Time-restricted” means that the attention mechanism is conducted on a limited window of frames to the left and right. In this study, we first explore time-restricted location-aware attention CTC/Attention, establishing the proper time-restricted attention window size. Inspired by the success of self-attention in machine translation, we further introduce the time-restricted self-attention CTC/Attention that can better model the long-range dependencies among the frames. Experiments with wall street journal (WSJ), augmented multiparty interaction (AMI), and switchboard (SWBD) tasks demonstrate the effectiveness of the proposed time-restricted self-attention CTC/Attention. Finally, to explore the robustness of this method to noise and reverberation, we join a train neural beamformer frontend with the time-restricted attention CTC/Attention ASR backend in the CHIME-4 dataset. The reduction of word error rate (WER) and the increase of perceptual evaluation of speech quality (PESQ) approve the effectiveness of this framework.

Common sound source localization algorithms focus on localizing all the active sources in the environment. While the source identities are generally unknown, retrieving the location of a speaker of interest requires extra effort. This paper addresses the problem of localizing a speaker of interest from a novel perspective by first performing time-frequency selection before localization. The speaker of interest, namely the target speaker, is assumed to be sparsely active in the signal spectra. The target speaker-dominant time-frequency regions are separated by a speaker-aware Long Short-Term Memory (LSTM) neural network, and they are sufficient to determine the Direction of Arrival (DoA) of the target speaker. Speaker-awareness is achieved by utilizing a short target utterance to adapt the hidden layer outputs of the neural network. The instantaneous DoA estimator is based on the probabilistic complex Watson Mixture Model (cWMM), and a weighted maximum likelihood estimation of the model parameters is accordingly derived. Simulative experiments show that the proposed algorithm works well in various noisy conditions and remains robust when the signal-to-noise ratio is low and when a competing speaker exists.

This paper proposes a note-based music language model (MLM) for improving note-level polyphonic piano transcription. The MLM is based on the recurrent structure, which could model the temporal correlations between notes in music sequences. To combine the outputs of the note-based MLM and acoustic model directly, an integrated architecture is adopted in this paper. We also propose an inference algorithm, in which the note-based MLM is used to predict notes at the blank onsets in the thresholding transcription results. The experimental results show that the proposed inference algorithm improves the performance of note-level transcription. We also observe that the combination of the restricted Boltzmann machine (RBM) and recurrent structure outperforms a single recurrent neural network (RNN) or long short-term memory network (LSTM) in modeling the high-dimensional note sequences. Among all the MLMs, LSTM-RBM helps the system yield the best results on all evaluation metrics regardless of the performance of acoustic models.

This paper presents a two-stage transcription framework for a specific piano, which combines deep learning and spectrogram factorization techniques. In the first stage, two convolutional neural networks (CNNs) are adopted to recognize the notes of the piano preliminarily, and note verification for the specific individual is conducted in the second stage. The note recognition stage is independent of piano individual, in which one CNN is used to detect onsets and another is used to estimate the probabilities of pitches at each detected onset. Hence, candidate pitches at candidate onsets are obtained in the first stage. During the note verification, templates for the specific piano are generated to model the attack of note per pitch. Then, the spectrogram of the segment around candidate onset is factorized using attack templates of candidate pitches. In this way, not only the pitches are picked up by note activations, but the onsets are revised. Experiments show that CNN outperforms other types of neural networks in both onset detection and pitch estimation, and the combination of two CNNs yields better performance than a single CNN in note recognition. We also observe that note verification further improves the performance of transcription. In the transcription of a specific piano, the proposed system achieves 82% on note-wise F-measure, which outperforms the state-of-the-art.

Laser based positron represents a new particle source with short pulse duration and high charge density. Positron production based on laser wakefield electron acceleration (LWFA) has been investigated theoretically in this paper. Analytical expressions for positron spectra and yield have been obtained through a combination of LWFA and cascade shower theories. The maximum positron yield and corresponding converter thickness have been optimized as a function of driven laser power. Under the optimal condition, high energy (>100 MeV) positron yield up to 5×10^{11} can be produced by high power femtosecond lasers at ELI-NP. The percentage of positrons shows that a quasineutral electron-positron jet can be generated by setting the converter thickness greater than 5 radiation lengths.

Robust automatic language identification (LID) is a task of identifying the language from a short utterance spoken by an unknown speaker. One of the mainstream approaches named parallel phone recognition language modeling (PPRLM) has achieved a very good performance. The log-likelihood radio (LLR) algorithm has been proposed recently to normalize posteriori probabilities which are the outputs of back-end classifiers in PPRLM systems. Support vector machine (SVM) with radial basis function (RBF) kernel is adopted as the back-end classifier. But for the conventional SVM classifier, the output is not probability. We use a pair-wise posterior probability estimation (PPPE) algorithm to calibrate the output of each classifier. The proposed approaches are evaluated on the 2005 National Institute of Standards and Technology (NIST). Language recognition evaluation databases and experiments show that the systems described in this paper produce comparable results to the existing arts.