The segmentation of liver and tumor images plays a supportive role in clinical diagnosis. Liver tumors vary significantly in size and exhibit diverse morphological variations. Additionally, liver tumors are similar to their surrounding tissues. These obstacles hinder algorithms from enhancing their effectiveness. In this paper, PVTv2 is employed as the encoder on the left side of the model. Multiscale information extraction modules are used at the skip connections to capture rich feature representations. To distinguish similar tissues, the model enhances uncertain areas by attention mechanism. The global dice coefficient for liver on the LiTS dataset reached 97.33%, with 82.24% for tumor. Our model outperforms the other models in comparison. Our ablation experiments validate the effectiveness of our modules.

This paper addresses the problem of vascular segmentation of retinal fundus images, which is prone to breakage and missing microvessel segmentation, from the perspective of enabling the network to acquire more abundant and significant semantic information. First, using U-Net as the backbone network, the novel residual attention module RSCA and multiscale transformer as generators, and the binary classification model with dense jump connections as discriminators, which are combined to form a multiscale GAN network. Second, setting the joint loss function for parameter update. The accuracy, sensitivity, specificity, and ACC of the constructed model on the DRIVE dataset reached 98.57%, 84.13%, 98.63%, and 97.80%, respectively, which is at the level of human experts.

This work suggests a medical image segmentation approach based on local-global Gaussian-weighted attention as a remedy for the U-Net model’s lack of explicit modeling of long-range dependencies for medical image segmentation problems. The approach uses a Transformer module built on a local-global Gaussian weighted attention mechanism and a unified decoder module in the U-Net model to enable relevant relationship mining and centralized multi-resolution feature processing. Additionally, this paper validates the efficacy of the method put forth in this paper on two openly accessible datasets, Synapse and ACDC, by expanding the dataset with random angle rotation and mirroring operations for pre-processing and using a loss function combining HD95 and Dice for training. According to the experimental findings, the suggested method is competitive with other existing methods and outperforms TransUNet and U-Net in terms of the model’s generalization and segmentation effects.

In recent years, using convolutional neural network (CNN) to segment medical images has achieved good results already. The CNN model with U-shaped structure of encoder and decoder and jump connection mechanism has been widely used in various medical tasks. However, CNN is unable to learn global information and conduct remote semantic information interaction due to its inductive bias. With the application of Transformer in Computer Vision field, the global remote dependency modeling capability brought by Transformer can make up for the locality of CNN to some extent. However, in recent studies, most proposed methods only use Transformer or combinations of Transformer and CNN at both encoder and decoder. In this paper, we propose a new U-shaped network framework with incomplete symmetry, using Transformer module for image feature extraction, and using CNN for image recovery, to seek more possibilities for the combination of Transformer and CNN in the medical segmentation field. Experiments on cardiac MRI segmentation ACDC data sets show that giving up Transformer in the decoding layer will not reduce the overall segmentation performance, and half-transformer and half-CNN network structure can give a good balance between computation and segmentation performance.

According to the characteristics of kidney tumours in abdominal multi-organ CT images, analyzing the CNN network to deal with long-distance connection defects and the shortage of Transformer structure detail processing, we designed a multimodal residual continuous convolution CNN and an improved multiscale dual-path multi-head attention Transformer combined with an encoder, deconvolution up-sampling, and learnable attention gating mechanism of the decoder with a jump junction introduced into the multiscale convolutional fusion module to form an asymmetric U-Net model framework for kidney tumour segmentation. Experimental comparison results show that the proposed method achieves an average dice coefficient of 0.9352, an average cross-merge rate of 0.9894, and a class-average pixel accuracy of 0.98235 for kidney segmentation metrics, which are more accurate and detailed than the existing and newer network models for segmentation.

With the improvement of medical consumption level, patients have more and more demand for the prediction of treatment costs. However, the prediction accuracy of existing research methods is low when the amount of data is small. In order to solve this problem, a weighted lasso regression method is proposed to predict the treatment cost based on the electronic medical record. Firstly, a set of transformation method of text-based medical record data is established, and the missing values are supplemented according to the clustering distance to realize the data representation of medical records. Then, in view of the low prediction accuracy of the traditional regression model, the lasso regression model with local weighting is established by introducing the data feature weight into lasso regression method. Finally, the model is verified by the medical record data provided by the hospital, and the results show that the model has higher prediction accuracy.

With the increasing development of smart grid technology, short-term load forecasting becomes particularly important in power system operation. However, the design of accurate and reliable short-term load forecasting methods and models is challenging due to the volatility and intermittency of renewable energy sources, as well as the privacy and individual characteristics of electricity consumption data from user data. To overcome this issue, in this paper, a novel cloud-edge collaboration short-term load forecasting method is proposed for smart grid. In order to reduce the computational load of edge nodes and improve the accuracy of node prediction, we use the method of building a model pre-training pool to train multiple pre-training models in the cloud layer at the same time. Then we use edge nodes to retrain the pre-trained model, select the optimal model and update the model parameters to achieve short-term load forecasting. To assure the validity of the model and the confidentiality of private data, we utilize the model pre-training pool to minimize edge node training difficulty and employ the approach of secondary edge node training. Finally, extensive experiments confirm the efficacy of our proposed method.

Power supply is the most basic guarantee for intelligent robots to complete complex instructions. In-depth research on power management technology has also accelerated the upgrading of robots. This paper designs a wide-input DC-DC power chip suitable for robot systems. To achieve the wide input of the chip, this article presents an improved LDO structure, which solves the problem that the traditional LDO structure cannot start normally under low voltage. In addition, this paper presents a new type of bootstrap capacitor charge and discharge control circuit, which solves the problem that the charging and discharging can't be completed quickly due to the low drop-out with traditional structure. Finally, in order to convert the unknown frequency information into a processable electrical signal, this paper presents a frequency conversion circuit that successfully converts the frequency information into a corresponding electrical signal. Through simulation verification, the power chip designed in this paper can work normally under a wide input voltage range of 1.6V-6.3V.

In order to solve the function error in the cross clock domain transmission of high-speed data in NoC,a NoC cross clock domain processing circuit structure is presented.For synchronous NoC with multiple voltage frequency islands,multiplexers and asynchronous circular FIFOs are combined to constitute the NoC cross clock domain processing circuit.The experiment results show that the algorithm and the circuit design can effectively reduce the influence of metastable state phenomenon,increase the NoC data transmission throughput,and satisfy the requirement of real-time transmission of high-speed data in NoC used for video acquisition and processing.

Sensors are used to collect the natural signals, including rarely change signals. Level-crossing ADC (LC-ADC) employs the idea of irregular sampling, which is adapted to convert the sparse signals. The comparator is an important block of the converter and hysteresis is useful for removing erroneous sampling due to noise. Two hysteresis comparators are presented, one is used in the detection of the signal down and the other is used in the detection of the signal up. The comparators have been implemented in 0.18μm CMOS technology, the area is 940.7μm 2 and 864.5μm 2 respectively, and the propagation delay for the low-to-high transition is 9.5ns and the propagation delay for the high-to-low transition is 10.5ns with the common mode input voltage is 800mV.

This study's aim was to investigate the post-effect of an air quality improvement on systemic inflammation and circulating microparticles in asthmatic patients during, and 2 months after, the Beijing Olympics 2008. We measured the levels of circulating inflammatory cytokines and microparticles in the peripheral blood from asthma patients and healthy controls during (phase 1), and 2 months after (phase 2) the Beijing 2008 Olympic Games. The concentrations of circulating cytokines (including TNFα, IL-6, IL-8, and IL-10) were still seen reduced in phase 2 when compared with those in phase 1. The number of circulating endothelial cell-derived microparticles was significantly lower during the phase 2 than that during phase 1 in asthma patients. The level of plasma lipopolysaccharide-binding protein (LBP) was significantly decreased in asthmatics in phase 2. The level of norepinephrine was significantly higher in phase 2 than that in phase 1 in plasma from both asthma patients and healthy subjects. There were no significant differences in the gene profile for the toll-like receptor (TLR) signaling from peripheral blood mononuclear cells. In vitro, microvesicles from patients with asthma impaired the relaxation to bradykinin and contraction to acetylcholine, whereas microparticles from healthy subjects did not. These data suggested that reduction in systemic pro-inflammatory responses and circulating LBP and increased level of norepinephrine in asthma patients persisted even after 2 months of the air pollution intervention. These changes were independent of the TLR signaling pathway. Circulating microparticles might be associated with airway smooth muscle dysfunction.

Realgar (As4S4), as a mineral drug containing arsenic compound, has been employed in clinical therapy of cancer for its good therapeutic reputation in Chinese traditional medicine. However, large dose of realgar and long period of treatment are necessary for achieving the effective blood medicine concentration due to its low bioavailability resulted from poor solubility. In this study, we obtained realgar transforming solution (RTS) using intrinsic biotransformation in microorganism, and investigated underlying mechanisms of RTS for HepG2 cells. Our results demonstrated that an effective biotransformation of realgar method by A. ferrooxidans was established, in which realgar was biologically converted into an aqueous solution, and RTS had a strong activity inducing apoptosis and interrupting G2/M progression in HepG2 cells via upregulation of cellular ROS. Importantly, RTS inhibited the cellular antioxidant defense system leading to abundant ROS accumulation, and activated cell cycle arrest and mitochondrial pathway of apoptosis mediated by activating p53 due to cellular uncontrolled ROS. Collectively, our findings suggest that RTS is a potential candidate for therapy of human hepatocellular carcinoma.

Air pollution is associated with the increased risk of metabolic syndrome. In this study, we performed inhalation exposure of mice fed normal chow or a high-fat diet to airborne fine particulate matters (PM2.5), and then investigated the complex effects and mechanisms of inhalation exposure to PM2.5 on hepatic steatosis, a precursor or manifestation of metabolic syndrome. Our studies demonstrated that inhalation exposure of mice fed normal chow to concentrated ambient PM2.5 repressed hepatic transcriptional regulators involved in fatty acid oxidation and lipolysis, and thus promoted hepatic steatosis. However, PM2.5 exposure relieved hepatic steatosis in high-fat diet-induced obese mice. Further investigation revealed that inhalation exposure to PM2.5 induced hepatic autophagy in mouse livers in a manner depending on the MyD88-mediated inflammatory pathway. The counteractive effect of PM2.5 exposure on high-fat diet-induced hepatic steatosis was mediated through PM2.5-induced hepatic autophagy. The findings from this study not only defined the effects and mechanisms of PM2.5 exposure in metabolic disorders, but also revealed the pleotrophic acts of an environmental stressor in a complex stress system relevant to public health.

Background Chronic exposure to fine ambient particulate matter (PM 2.5 ) induces insulin resistance. CC-chemokine receptor 2 (CCR2) appears to be essential in diet-induced insulin resistance implicating an important role for systemic cellular inflammation in the process. We have previously suggested that CCR2 is important in PM 2.5 exposure-mediated inflammation leading to insulin resistance under high fat diet situation. The present study assessed the importance of CCR2 in PM 2.5 exposure-induced insulin resistance in the context of normal diet. Methods and Results C57BL/6 and CCR2 -/- mice were subjected to exposure to concentrated ambient PM 2.5 or filtered air for 6 months. In C57BL/6 mice, concentrated ambient PM 2.5 exposure induced whole-body insulin resistance, macrophage infiltration into the adipose tissue, and upregulation of phosphoenolpyruvate carboxykinase (PEPCK) in the liver. While CCR2 deficiency reduced adipose macrophage content in the PM 2.5 -exposed animals, it did not improve systemic insulin resistance. This lack of improvement in insulin resistance was paralleled by increased hepatic expression of genes in PEPCK and inflammation. Conclusion CCR2 deletion failed to attenuate PM 2.5 exposure-induced insulin resistance in mice fed on normal diet. The present study indicates that PM 2.5 may dysregulate glucose metabolism directly without exerting proinflammatory effects.

The topology and routing algorithm of network-on-chip (NoC)directly influence the transmission delay and transmission efficiency of the network. A new topology of NoC— H-annular Mesh was proposed based on the 2D-Mesh topology. The lines introduced from the vertex nodes to the center nodes constituted a half annular Mesh (H-annular Mesh),which could fully take the advantage of 2D-Torus topology. An adaptive routing algorithm HAA-XY was proposed for H-annular Mesh topology. The simulation results indicate that based on the H-annular Mesh topology and the HAA-XY routing algorithm, the NoC can effectively reduce the network transmission delay, and can realize the multi-directional and multi-node data parallel communication.

In order to improve the robustness of Real-time registration of augmented reality,this paper proposed IFREAK feature descriptor. It was proposed for the characteristic of FREAK feature descriptor. The descriptor was based on sampling pattern of FREAK descriptors and it was considered the spatial structure. In IFREAK descriptor,every receptive field pair stored multiple bits,and the number of the bits depended on the parameters. Compared with FREAK descriptor storage,IFREAK descriptor improves the accuracy of feature points matching. Based on FREAK feature descriptor,the paper proposed an improved real-time registration methods of augmented reality. Experimental results showthat,in most scenes,matching of the proposed descriptor is better than matching of FREAK descriptor and others. The proposed algorithm has stronger robustness.

The high erroneous results of the stereo matching occur at the following three cases where there are depth discontinuity region,the slanted surface or the non-fronto-parallel surface.A stereo matching algorithm was proposed based on the improved Patchmatch and slice sampling particle belief propagation. An edge-preserving similarity function of the Patchmatch was defined.Then,a model of the depth estimation for the non-fronto-parallel surface was introduced. The nearest neighbor search was replaced with the particle belief propagation, and the target distribution was approximated with a finite set of particles. At the same time,the sampled particles from the belief distribution was typically done by using slice sampling Markov chain Monte Carlo method to solve the particle update problem. The experiments on the Middlebury indicate that the mismatching at the depth discontinuity region can be reduced,and the match accuracy for the slanted surface and the non-fronto-parallel surface can be improved.

针对传统D-S证据理论难以解决高度冲突证据融合问题,提出一种新的证据合成算法.将贴近度概念引入D-S证据合成中,通过证据的一致性度量计算其权重,实现冲突证据的加权融合.提出证据合成方法选择判据,将证据合成分为冲突和非冲突2类,分别采用改进算法和传统算法对证据进行融合.实例验证表明,所提出的方法信息聚焦性能优越,可以有效解决冲突证据合成问题,在解决电力系统故障诊断问题方面有良好的效果.

This paper proposes a single track two phase static asynchronous Pipeline buffer, circuit design based on the 1-of-N handshake protocol to communicate, the circuit can complete data transmissions properly without acknowledgement signals, It is a kind of dual-rail buffer and provides a tradeoff between area and performance, using domino logic, effectively improve the utilization rate of the area and has very good latency characteristics. Simulations are made based on TSMC 0.18μm CMOS technology, the results present that the circuit can work up to 2.65 GHz under ten stages in series, and the forward transmission delay is very low for 116 ps circuit, superior to the STFB(single-track full-buffer) and Gas P and circuit, the results presents that the circuit suitable for medium- to high-performance asynchronous circuit design.

iming at the problem that nonlinear estimation results of Extended Kalman Filter-based Simultaneous Localization and Mapping(EKF-SLAM) algorithm are inconsistent,this paper proposes an Improved Extended Kalman Filter-based Simultaneous Localization and Mapping(IEKF-SLAM) algorithm with polynomial.And on this basis,it designs a tracking and registration algorithm of Augmented Reality(AR) on unknown scene including mapping and updating,tracking and registration two parallel modules.Mapping and updating module uses the IEKF-SLAM algorithm.The tracking and registration module after video frame is captured,camera pose is estimated by constructing a map library.Then video frames feature points are extracted and the feature points are matched to the map library.The pose of the camera is updated.Then virtual objects are rendered and registered.Experimental results show that the consistency of the IEKF-SLAM algorithm is superior to the EKF-SLAM algorithm,and the result of tracking and registration of AR is satisfactory.