Search

Crowd counting is getting more and more attention in our daily life, because it can effectively prevent some safety problems. However, due to scale variations and background noise in the image, such as buildings and trees, getting the accurate number from image is a hard work. In order to address these problems, this work introduces a new multi-scale supervised network. The proposed model uses part of vgg16 model as the backbone to extract feature. In the training process, a multi-scale dilated convolution module is added at the end of each stage of the backbone network to generate attention map with different resolutions to help the model focus on the head area in feature map. In addition, the dilated convolution adopts three dilation ratios to fit different sizes of head in the image. Finally, in order to get the high-quality density map with high-resolution, the authors employ the upsampling operation to restore the density map size to the quarter size of original image. A large number of experiments on these four datasets show that the proposed network has greatly improved the counting accuracy of many existing methods.

To summarize the strategy of using extracorporeal membrane oxygenation (ECMO) support during lung transplantation from 2 coronavirus disease 2019 (COVID-19) with end-stage respiratory failure.Two COVID-19 with end-stage respiratory failure patients were admitted to Nanjing Medical University Affiliated Wuxi People's Hospital in March 2020. As the homoeostasis and vital signs could not be maintained in balance by conventional treatments, lung transplantations were performed. Here, detail information about combined application of peripheral veno-venous ECMO (VV-ECMO) and central veno-arterial ECMO (CVA-ECMO) during the operation will be discussed.Case 1: 59 years old, 172 cm height, 72 kg weight, who received mechanical ventilation for 22 days, tracheotomy tube for 17 days, and VV-ECMO support for 7 days. Case 2: 72 years old, 178 cm height, 71 kg weight, who received mechanical ventilation for 19 days, tracheotomy tube for 17 days, and VV-ECMO support for 18 days. As both of them have severe COVID-19-associated respiratory failure, and the recovery was determined to be unlikely, lung transplantations were performed. Severe pulmonary arterial hypertension (PAH) and cardiac insufficiency were found during the operation. Based on preoperative VV-ECMO, CVA-ECMO was added. The concomitant use of peripheral VV-ECMO and CVA-ECMO offered satisfied intraoperative oxygenation and cardiopulmonary status,the operations run smoothly, and the CVA-ECMO was successfully removed, no ECMO-related complications occurred.The combined use of VV-ECMO and CVA-ECMO is an optimal strategy in the end-stage ARDS patients with severe PAH and cardiac insufficiency, which can offer benefits on respiratory and cardiac functions simultaneously, and ensure surgery safety.

During the wireless charging of an automated guided vehicle (AGV), the output voltage is unstable due to changes in parameters such as coil mutual inductance and load resistance caused by external interferences and internal mismatches of the system. In this paper, an integral sliding mode control method based on an unknown input observer (UIO) containing predictive equations is designed to build an inductor–capacitor–capacitor-series (LCC-S) topology model for wireless power transfer (WPT). The observer designed by this method can perceive changes in the secondary resistance parameter and the mutual inductance of the primary and secondary coils. The design with the prediction equation speeds up the convergence of the observer to the true value. The observer’s compensation of the control system avoids the occurrence of integral oversaturation. The experimental results show that, based on the UIO-SMC system output, voltage can be accurately controlled to meet the requirement for a given voltage. The effect of suppressing disturbance is better than with SMC and PI control. When the system parameter changes, it has better voltage anti-interference performance and stronger ripple suppression.

Objective: Despite the application of dexmedetomidine (DEX) as a perioperative adjuvant in local analgesia, the exact analgesic mechanism underpinning chronic neuropathic pain (CNP) awaits our elucidation. Methods: We investigated the molecular mechanisms of the anti-nociceptive effect of DEX on neuropathic pain in a mouse model induced by chronic constriction injury (CCI). Results: DEX administration significantly increased the paw withdrawal latency (PWL) values 0.5 to 2 h post-injection in CCI-induced CNP mice at day 5 to 21 versus dimethyl sulfoxide (DMSO)-treated mice, confirming its analgesic effect. The c-Fos expression was significantly elevated in CCI mice versus the sham-operated group, whereas the elevation was mitigated by DEX injection. Subsequently, the involvement of MKP1 and MKP3 in the pathogenesis of chronic neuropathic pain was evaluated. Western blotting analyses revealed significant decrease in both MKP1 and MKP3 in the spinal cord in CCI group versus the sham group. DEX markedly elevated the MKP3 expression and modestly reduced the MKP1 expression, with insignificant difference in the latter. Co-injection of BCI (an MKP3 inhibitor) and DEX evidently reduced the PWL values in CCI mice. Furthermore, DEX significantly downregulated the phosphorylation of extracellular-signal-regulated kinase (ERK) 1/2, down-stream effector of MKP3 in CCI mice, whereas the downregulation was reversed by BCI. Conclusion: We confirmed that DEX exerts the analgesic effect on chronic neuropathic pain via the regulation of MKP3/ERK1/2 signaling pathway, which may contribute to clarification of the molecular mechanism and novel therapy for chronic neuropathic pain.

In order to improve dynamic performance and robustness of interleaved boost converter, this paper proposes a predictive control strategy based on deadbeat method using Luenberger disturbance observer (LDO). To solve the problem of modeling inaccuracy under the condition of input voltage fluctuation and unknown inductor equivalent series resistance, a LDO is designed to observe the lumped disturbance and compensate the predictive model. Subsequently, another LDO is proposed to estimate load equivalent resistance to enhance the robustness under load perturbation condition. Furthermore, the carrier phase shifting pulse width modulation (CPS-PWM) is used to reduce output current ripple combining with interleaved topology. The simulation results verify the excellent dynamic performance and robustness of proposed method.

BACKGROUND: Lactic acidosis is often seen in lung transplantation (LTx). Postoperative lactate is frequently associated with poor outcome in postoperative and critically ill patients. Our aim was to evaluate the predictive value of postoperative peak lactate levels within 72 h of LTx for 30-day and late mortality. METHODS: We evaluated patients who underwent LTx from January 2015 to September 2017. All admitted patients were classified according to the peak lactate level (PL) within 72 h of surgery: PL 10 mmol/L (Group 3). We performed logistic regression analysis and used Cox regression models to identify the peak lactate level as a predictive factor for 30-day and late mortality, respectively. RESULTS: Of 255 eligible patients, mean age 55.61±12.16, mean lactate 4.99±2.93 and 80% male, and 40% had hyperlactatemia (PL >5 mmol/L) after LTx. The 30-day mortality rate was 17.9%, 28.9% and 68.8% in the three groups, respectively (P

With the advent of the big data era, knowledge graph embodies great advantages. Especially in game domain, building a knowledge graph is of great value. However, some open-domain knowledge bases is not designed for games and the amount of game data is relatively limited. In this paper, we propose a game ontology. In addition, in the process of constructing game knowledge graph, we found it is import to identify an entity is a game entity or not. Thus, we propose a new DLC entity identification model to help us use the ontology to build the graph.

Background Studies have provided some evidence that pain is a risk factor for postoperative delirium (POD). Therefore, we investigated the relationship between preoperative pain and POD after noncardiac surgery. Methods POD was assessed with the Montreal Cognitive Assessment, and preoperative cognition was assessed with the Mini-Mental State Examination. Plasma C-reactive protein (CRP) was detected by enzyme-linked immunosorbent assay before surgery. Preoperative pain was classified by its duration before surgery as chronic pain (lasting more than 1 month), acute pain (lasting less than 1 month), or no pain (no obvious pain). Multiple linear regression was used to adjust for confounding. Results From October 15, 2018, through August 12, 2019, a total of 67 patients were randomized; 7 were excluded because they were discharged before the seventh postoperative day. The prevalence of POD was significantly higher in the acute pain group (13 of 20; 65%) than in the chronic pain group (5 of 20; 25%) or the no pain group (6 of 20; 30%) (P = 0.019), indicating that delirium is associated with preoperative acute pain. The plasma level of preoperative CRP was also higher in the acute pain group than in the other two groups (mean [interquartile range]: 10.7 [3.3, 29.3] vs 1 [0.5, 3.8]mg/l; P Conclusions Preoperative acute pain was associated with POD, and increased plasma levels of CRP provide a marker. In addition, we found that illiteracy and advanced age were risk factors for POD.

Aiming at the bottleneck of thermodynamics, kinetics and reaction systems in photothermal catalytic CO2 reduction conversion, a three-field coupling method based on the efficient utilization of solar energy and the application of a photo-thermal-magnetic three-field was proposed to achieve efficient and highly selective CO2 conversion. The composite monolithic catalyst with three field response functions and multiple active reaction interfaces was prepared by using photoactive copper supported by nickel foam. The effect of an external alternating magnetic field on photocatalytic CO2 reduction was studied using a range between 0.005 and 0.1 wt% Cu/Cu2O/Ni(OH)2/NF as catalysts. The results showed that 0.01 wt% Cu/Cu2O/Ni(OH)2/NF exhibited the best catalytic activity and CH4 selectivity under alternating magnetic field enhanced photocatalysis. The main products were CO (6.76 μmol g−1) and CH4 (167 μmol g−1). The selectivity of CH4 was 96.1%. The yields of CH4 using this method were 11 and 6 times higher than the yields obtained with photocatalysis (14.58 μmol g−1) and magnetic-thermal catalysis (26.75 μmol g−1), respectively. According to the analysis of temperature measurements and electrochemical measurements, the surface temperature of the three-field coupled catalyst reaches about 230 °C. Applying an alternating magnetic field can effectively reduce the recombination of photogenerated electron holes. Alternating magnetic field enhanced photocatalytic CO2 reduction conversion provides a new and promising method for the efficient conversion and utilization of solar energy.

In wireless power transfer systems, mutual inductance and load parameters play important roles. They affect the efficiency and output ability of the system and often change in real situations. Therefore, in order to achieve better real-time control of the system and improve its dynamic performance, it is necessary to identify these parameters. In this paper, a new parameter identification algorithm is proposed. This algorithm combines the model predictive control algorithm with the particle swarm optimization algorithm to achieve the desired control target based on the principle of the magnetic coupling resonance. The obtained control signal is input into the pulse density modulation to realize the control of the inverter and rectifier. Compared with the original algorithm, it has two distinct characteristics. One is that its dynamic performance is better and its scope of work is wider, and the other is that it has a simple hardware structure and is easier to implement. The simulation are carried out in order to verify the feasibility of the proposed algorithm.

Cardiopulmonary bypass (CPB) can induce inflammatory lung injury, which is a common complication during cardiac surgery. Nod-like receptor family, pyrin domain-containing 3 (NLRP3) inflammasome-induced inflammation plays a crucial role in lung injury after CPB. Previous studies have shown that electroacupuncture (EA) has potential anti-inflammatory activity. However, the role of EA in CPB is poorly understood. The aim of this study was to determine whether EA was associated with CPB-induced inflammatory lung injury. In the present study, rats were treated with EA for 5 days before CPB. Two hours after CPB, the lung tissue, serum, and bronchoalveolar lavage fluid (BALF) were prepared for assessment. Our results showed that the expression of NLRP3 in the lung tissue increased significantly after CPB. The EA pretreatment suppressed NLRP3 inflammasome activation, reduced lung edema, and inhibited IL-1β release into the serum and BALF after CPB. Our findings suggest that EA pretreatment attenuates inflammatory lung injury after CPB by suppressing NLRP3 inflammasome activation.

With the rise of deep learning technology, person re-identification (Re-id) technology has been developed rapidly. During the training process, many recent methods are susceptible to target misalignment and without sufficient discriminative features. Aiming at these two problems, a simple and potent model is proposed by us. A new self-attention module and a multi-loss function with relative weight are designed to integrate the multi-level features of pedestrians in our network. Specifically, the goal of the self-attention is to instruct the baseline network to learn robust features from the resized images. The key of the self-attention model is the weighting of the importance of different person regions. Therefore, the non-local features of the different levels in the baseline network will be paid more attention, which is conducive to learn discriminative features from the baseline network. Finally, a multi-loss function with relative weight is introduced to enhance the feature learning ability and integrate more features reasonably. Many experiments have been done on the three datasets (Market1501, DukeMTMC-reID and CUHK03-NP) and the results explain that the new model gets a higher accuracy than many other recent approaches.

China maintains a leading position in egg production in the world, but the rate of the mass and the industrialized cleaning for eggs remain low. In this study, different preservation conditions and cleaning processes were applied on fresh eggs, a day after it has been laid. Some conditions were compared, including the coating process (single-layer coating and double–layer coating)—2 % polyvinyl alcohol (PVA) solution and 2‰ dehydrogenation sodium acetate solution, temperature of the solution, hot air drying, and white mineral oil coating. An analysis was conducted on the results, including the weight loss rate, Haugh value, pH value, height of albumen, and the color of yolk. The results are as follows: the color of yolk had no significant difference; set 7 had the highest Haugh value; sets 5, 6, 8 had the lowest weight loss rate; set 1 owned the highest pH value while set 8 the lowest. Heat treatment is useful for reducing the weight loss rate. Coating is good for reducing weight loss rate and delaying the decline of Haugh value.

Since wind turbine is a complex nonlinear and strong coupling system, traditional PI control method can hardly achieve good control performance. A self-optimizing pitch control method based on the active-disturbance-rejection control theory is proposed in this paper. A linear model of the wind turbine is derived by linearizing the aerodynamic torque equation and the dynamic response of wind turbine is transformed into a first-order linear system. An expert system is designed to optimize the amplification coefficient according to the pitch rate and the speed deviation. The purpose of the proposed control method is to regulate the amplification coefficient automatically and keep the variations of pitch rate and rotor speed in proper ranges. Simulation results show that the proposed pitch control method has the ability to modify the amplification coefficient effectively, when it is not suitable, and keep the variations of pitch rate and rotor speed in proper ranges

The principle and structure of a new self-scanning laser interferometer are introduced. The output of a self-scanning photodiode array is modulated to generate an ac measurement signal, thereby avoiding dc drift and low-frequency disturbances. Experiments show that the device can be used in the range of 100 mm and with a resolution of less than 0.02µm.