Multi-scale feature fusion and class weight loss for skin lesion classification.

Skin cancer is one of the common types of cancer. It spreads quickly and is not easy to detect in the early stages, posing a major threat to human health. In recent years, deep learning methods have attracted widespread attention for skin cancer detection in dermoscopic images. However, training a practical classifier becomes highly challenging due to inter-class similarity and intra-class variation in skin lesion images. To address these problems, we propose a multi-scale fusion structure that combines shallow and deep features for more accurate classification. Simultaneously, we implement three approaches to the problem of class imbalance: class weighting, label smoothing, and resampling. In addition, the HAM10000_RE dataset strips out hair features to demonstrate the role of hair features in the classification process. We demonstrate that the region of interest is the most critical classification feature for the HAM10000_SE dataset, which segments lesion regions. We evaluated the effectiveness of our model using the HAM10000 and ISIC2019 dataset. The results showed that this method performed well in dermoscopic classification tasks, with ACC and AUC of 94.0% and 99.3%, on the HAM10000 dataset and ACC of 89.8% for the ISIC2019 dataset. The overall performance of our model is excellent in comparison to state-of-the-art models.

Treatment of rectal anastomotic atresia with transurethral prostate resection instrumentation: A report of three cases.

Dixon surgery for rectal cancer can lead to severe intestinal narrowing and blockage that is difficult to treat with open surgery or colonoscopy. The aim of the present study was to develop a minimally invasive approach for treating rectal anastomotic atresia based on three cases that were managed with transurethral prostate resection instrumentation. Preoperative imaging determined the distance from the anastomotic closure to the anal margin, the length of the anastomotic closure and the degree of proximal intestinal dilation for all cases. During the procedure, the anastomotic site was visualized, and a circular electrode was used to excavate and open the blockage. Membrane-like closures were directly incised to achieve satisfactory results, with an anastomotic diameter >20 mm. Those cases with tubular atresia required an initial incision using the prostate resectoscope to relieve the obstruction, followed by radial incisions until achieving an anastomotic diameter >20 mm. At 3-6 months post-dilation, two of the patients with anastomotic atresia >20 mm had satisfactory bowel movements, whereas the remaining patient experienced tumor recurrence at the anastomotic site and discontinued treatment. This case series demonstrates the potential of transurethral prostate resection instrumentation as a safe and effective minimally invasive approach for rectal anastomotic atresia. Given that prostate resection instrumentation is readily available in hospitals in China, this approach is widely accessible to most patients. Furthermore, the technique leverages existing surgical technology and practices, requiring only a shift in the surgical site.

Conv-Swinformer: Integration of CNN and shift window attention for Alzheimer's disease classification.

Deep learning (DL) algorithms based on brain MRI images have achieved great success in the prediction of Alzheimer's disease (AD), with classification accuracy exceeding even that of the most experienced clinical experts. As a novel feature fusion method, Transformer has achieved excellent performance in many computer vision tasks, which also greatly promotes the application of Transformer in medical images. However, when Transformer is used for 3D MRI image feature fusion, existing DL models treat the input local features equally, which is inconsistent with the fact that adjacent voxels have stronger semantic connections than spatially distant voxels. In addition, due to the relatively small size of the dataset for medical images, it is difficult to capture local lesion features in limited iterative training by treating all input features equally. This paper proposes a deep learning model Conv-Swinformer that focuses on extracting and integrating local fine-grained features. Conv-Swinformer consists of a CNN module and a Transformer encoder module. The CNN module summarizes the planar features of the MRI slices, and the Transformer module establishes semantic connections in 3D space for these planar features. By introducing the shift window attention mechanism in the Transformer encoder, the attention is focused on a small spatial area of the MRI image, which effectively reduces unnecessary background semantic information and enables the model to capture local features more accurately. In addition, the layer-by-layer enlarged attention window can further integrate local fine-grained features, thus enhancing the model's attention ability. Compared with DL algorithms that indiscriminately fuse local features of MRI images, Conv-Swinformer can fine-grained extract local lesion features, thus achieving better classification results.

3D lamellar skeletal network of porous carbon derived from hull of water chestnut with excellent microwave absorption properties.

Biomass derived carbon has attracted extensive attention in the field of microwave absorption because of its sustainability and porous structure beneficial to microwave attenuation. In this study, 3D lamellar skeletal network porous carbon was successfully obtained from hull of water chestnut using biomass waste as raw material by controlling the ratio of KOH and precursors in a one-step carbonization process. The optimization of biomass carbon morphology was achieved and its microwave absorption properties were investigated. At the temperature of 600 °C, when the ratio of hull of water chestnut to KOH is 1:1, the porous carbon material with filling ratio of 35% can reach the effective absorption bandwidth (RL < -10 dB) of 6.0 GHz (12-18 GHz) at the matching thickness of 1.90 mm, covering the whole Ku band. When the thickness is 2.97 mm, the optimal reflection loss reaches -60.76 dB. The surface defects, interface polarization and dipole polarization of 3D porous skeleton network structure derived from hull of water chestnut contribute to the excellent reflection loss and bandwidth of porous carbon materials. The porous carbon with low density, low cost and simple preparation method has broad application prospects in the preparation of biomass-derived microwave absorbers.

VGG-TSwinformer: Transformer-based deep learning model for early Alzheimer's disease prediction.

BACKGROUND AND OBJECTIVE: Mild cognitive impairment (MCI) is a transitional state between normal aging and Alzheimer's disease (AD), and accurately predicting the progression trend of MCI is critical to the early prevention and treatment of AD. Brain structural magnetic resonance imaging (sMRI), as one of the most important biomarkers for the diagnosis of AD, has been applied in various deep learning models. However, due to the inherent disadvantage of deep learning in dealing with longitudinal medical image data, few applications of deep learning for longitudinal analysis of MCI, and the majority of existing deep learning algorithms for MCI progress prediction rely on the analysis of the sMRI images collected at a single time-point, ignoring the progressive nature of the disorder. METHODS: In this work, we propose a VGG-TSwinformer model based on convolutional neural network (CNN) and Transformer for short-term longitudinal study of MCI. In this model, VGG-16 based CNN is used to extract low-level spatial features of longitudinal sMRI images and map these low-level features to high-level feature representations, sliding-window attention is used for fine-grained fusion of spatially adjacent feature representations, and gradually fuses distant spatial feature representations through the superposition of attention windows of different sizes, temporal attention is used to measure the evolution of this feature representations as a result of disease progression. RESULTS: We validated our model on the ADNI dataset. For the classification task of sMCI vs pMCI, accuracy, sensitivity, specificity and AUC reached 77.2%, 79.97%, 71.59% and 0.8153 respectively. Compared with other cross-sectional studies also applied to sMRI, the proposed model achieved better results in terms of accuracy, sensitivity, and AUC. CONCLUSION: The proposed VGG-TSwinformer is a deep learning model for short-term longitudinal study of MCI, which can build brain atrophy progression model from longitudinal sMRI images, and improve diagnostic efficiency compared to algorithms using only cross-sectional sMRI images.

Application of Transurethral Prostate Resection Instrumentation for Treating Low Rectal Anastomotic Leakage: A Pilot Study.

Purpose: To determine an accurate method of inspecting low anastomotic leakages and application of transurethral prostate resection instrumentation for treating low rectal anastomotic leakage. Patients and Methods: Clinical data of eight patients treated for anastomotic leakage after rectal cancer surgery at Zhangye People's Hospital (affiliated to Hexi University), from August 2019 to November 2021, were retrospectively analyzed. Transanal prostate resection instrumentation was used to assess the leakage and surrounding conditions. Using prostate resection instrumentation, the presacral and perirectal residual cavities were washed and removed, and indwelling suprapubic presacral, transanal presacral, and rectal drainage tubes were placed. Continuous presacral saline irrigation and drainage and open negative-pressure suction in the rectal cavity were performed until the patients' fistula healed. Results: Of the eight patients with anastomotic leakages, one had grade B and seven had grade C International Study Group of Rectal Cancer anastomotic leakage classifications following Dixon operation. Transanal prostate resection instrumentation showed that the leakage of the one patient with grade B was less than a third of the circumference of the anastomosis. Among the seven patients with grade C, one leakage was less than a third of the anastomotic circumference. One patient had complete separation of the anastomosis and one distal colon necrosis, which necessitated immediate descending colostomy. Conservative treatment was successful in six patients; the conservative overall cure rate was 75%, and the median healing time was 43 (21-68) days. Conclusion: Transanal examination of rectal anastomotic leakage using prostate resection instrumentation is comprehensive, easy to perform, provides clear visualization, accurately guides catheter placement, and can be combined with continuous open negative-pressure drainage, which is a safe, convenient, and effective method for treating low rectal leakage.

Highly efficient solar photothermal conversion of graphene-coated conjugated microporous polymers hollow spheres.

Highly efficient harvesting, transfer, and storage of solar energy are of great significance for the sustainability of society Herein, we report the design and synthesis of conjugated microporous polymers hollow spheres (CMPs-HS) coated by graphene (GCMPs-HS) and compounded with the phase change material (PCM) octadecanol (GCMPs@ODA) for efficient solar photothermal conversion. The as-synthesized CMPs-HS shows a high specific surface (519.95 m2 g-1 and 309.26 m2 g-1), good thermostability, and lower thermal conductivity (0.33 W m-2h-1). By coating graphene, the light absorption remained about 90% in the visible light range, which allows light harvest for photothermal conversion. Taking the GCMPs-HS as a functional layer for the solar steam generation (SSG) system, a high evaporation efficiency of near 90% is obtained. After inhaling octadecanol, GCMPs@ODA are prepared and their latent heats are measured to about 217.4 J g-1 and 224.6 J g-1. Under 1 sun irradiation, the photothermal conversion efficiencies of GCMPs@ODA are measured to be 87.15% and 85.83%, the above merits applied in different conditions are superior to photothermal conversion materials reported in the literature. Thus, among the above merits, the fabricated materials are the competitive candidate which shows the great potential in the efficient application of solar energy.

Robust Security Beamforming for SWIPT-Assisted Relay System with Channel Uncertainty.

This paper considers the physical layer security (PLS) of a simultaneous wireless information and power transfer (SWIPT) relay communication system composed of a legitimate source-destination pair and some eavesdroppers. Supposing a disturbance of channel status information (CSI) between relay and eavesdroppers in a bounded ellipse, we intend to design a robust beamformer to maximum security rate in the worst case on the constraints of relay energy consumption. To handle this non-convex optimization problem, we introduce a slack variable to transform the original problem into two sub-problems firstly, then an algorithm employing a semidefinite relaxation (SDR) technique and S-procedure is proposed to tackle above two sub-problems. Although our study was conducted in the scene of a direct link among source, destination, and eavesdroppers that is non-existing, we demonstrate that our conclusions can be easily extended to the scene for which a direct link among source, destination and eavesdroppers exist. Numerical simulation results compared with the benchmark scheme are provided to prove the effectiveness and superior performance of our algorithm.

Superwetting Electrospun PDMS/PMMA Membrane for PM2.5 Capture and Microdroplet Transfer.

Efficient acquiring and removal of a hazardous particulate matter (PM) have significant effects on human health. Here, we illustrate the fabrication of a superwetting electrospun polydimethylsiloxane/polymethyl methacrylate (PDMS/PMMA) membrane (EPPM) with multifunctional performance for PM2.5 capture and microdroplet transfer, where PMMA was added as a carrier polymer to the superhydrophobic PDMS, which has very low cohesive energy density. The obtained EPPM, which is composed of special bead-on-string fibers with a mean fiber diameter of 350 nm, shows a porous structure with an aperture of 7.87 µm (calculated by the bubble pressure method) and superb thermostability (up to 325 °C). The EPPM possesses an excellent PM2.5 purification efficiency of nearly up to 100% at a very low pressure drop (70 Pa, <0.07% of the atmospheric pressure) under the condition of high humidity (96 ± 3%), which is greatly advantageous over those hydrophilic filters frequently suffering the drawbacks of low efficiency or total invalidation in humid environments. In addition, benefitting from the superhydrophobic and strong adhesive properties of the membrane surface, the EPPM could complete the trace aqueous sample analysis such as "robotic hand" from superhydrophobic to hydrophilic surfaces without any contamination or loss and hold a high contact angle of 161.6° for water. Altogether, the EPPM may have technological advantages as a kind of novel fibrous filter in diverse environmental applications, including PM2.5 capture, separation, microdroplet transfer, and so on.

A Cluster-Based Energy Optimization Algorithm in Wireless Sensor Networks with Mobile Sink.

Aiming at high network energy consumption and data delay induced by mobile sink in wireless sensor networks (WSNs), this paper proposes a cluster-based energy optimization algorithm called Cluster-Based Energy Optimization with Mobile Sink (CEOMS). CEOMS algorithm constructs the energy density function of network nodes firstly and then assigns sensor nodes with higher remaining energy as cluster heads according to energy density function. Meanwhile, the directivity motion performance function of mobile sink is constructed to enhance the probability of remote sensor nodes being assigned as cluster heads. Secondly, based on Low Energy Adaptive Clustering Hierarchy Protocol (LEACH) architecture, the energy density function and the motion performance function are introduced into the cluster head selection process to avoid random assignment of cluster head. Finally, an adaptive adjustment function is designed to improve the adaptability of cluster head selection by percentage of network nodes death and the density of all surviving nodes of the entire network. The simulation results show that the proposed CEOMS algorithm improves the cluster head selection self-adaptability, extends the network life, reduces the data delay, and balances the network load.

Strong Tracking PHD Filter Based on Variational Bayesian with Inaccurate Process and Measurement Noise Covariance.

Assuming that the measurement and process noise covariances are known, the probability hypothesis density (PHD) filter is effective in real-time multi-target tracking; however, noise covariance is often unknown and time-varying for an actual scene. To solve this problem, a strong tracking PHD filter based on Variational Bayes (VB) approximation is proposed in this paper. The measurement noise covariance is described in the linear system by the inverse Wishart (IW) distribution. Then, the fading factor in the strong tracking principle uses the optimal measurement noise covariance at the previous moment to control the state prediction covariance in real-time. The Gaussian IW (GIW) joint distribution adopts the VB approximation to jointly return the measurement noise covariance and the target state covariance. The simulation results show that, compared with the traditional Gaussian mixture PHD (GM-PHD) and the VB-adaptive PHD, the proposed algorithm has higher tracking accuracy and stronger robustness in a more reasonable calculation time.

Salt-Rejection Solar Absorbers Based on Porous Ionic Polymers Nanowires for Desalination.

The construction of photothermal materials with ideal salt tolerance has been a major subject for efficient solar desalination. Herein, a novel photothermal material based on porous ionic polymers (PIPs) nanowires is synthesized by Sonogashira-Hagihara cross-coupling reaction using ionic salt and alkynylbenzene as building blocks. The PIPs nanowires monolith shows abundant porosity with low density, leading a superior thermal insulation. The intrinsic superhydrophilicity of PIPs nanowires endows it with desired water transportation ability. By facile spraying Chinese carbon-ink on the PIPs nanowires monolith, its light absorption can be enhanced to be 90%. Based on these merits, the PIPs nanowires based photothermal materials show high solar energy conversion efficiency (81% under 1 sun irradiation). More interestingly, its inherently ionic framework can result in an ion-ion interaction between the external ions in water and ionic groups in PIPs framework, thus leading to excellent desalination ability by combing its unique superhydrophilicity, for example, no salt accumulation is observed after 6 h duration at 1 sun irradiation. Compared with the existing salt-resistant photothermal materials, the method takes the advantage of the intrinsically ionic feature of PIPs without using any artificial process, thus may open a new way for design and fabrication of high-performance salt-rejection photothermal materials.

Robust Visual Tracking Based on Adaptive Multi-Feature Fusion Using the Tracking Reliability Criterion.

Multi-resolution feature fusion DCF (Discriminative Correlation Filter) methods have significantly advanced the object tracking performance. However, careless choice and fusion of sample features make the algorithm susceptible to interference, leading to tracking failure. Some trackers embed the re-detection module to remedy tracking failures, yet distinguishing ability and stability of the sample features are scarcely considered when training the detector, resulting in low effectiveness detection. Firstly, this paper proposes a criterion of feature tracking reliability and conduct a novel feature adaptive fusion framework. The feature tracking reliability criterion is proposed to evaluate the robustness and distinguishing ability of the sample features. Secondly, a re-detection module is proposed to further avoid tracking failures and increase the accuracy of target re-detection. The re-detection module consists of multiple SVM detectors trained by different sample features. When the tracking fails, the SVM detector trained by the most reliable sample feature will be activated to recover the target and adjust the target position. Finally, comparison experiments on OTB2015 and UAV123 databases demonstrate the accuracy and robustness of the proposed method.

Observable Degree Analysis for Multi-Sensor Fusion System.

Multi-sensor fusion system has many advantages, such as reduce error and improve filtering accuracy. The observability of the system state is an important index to test the convergence accuracy and speed of the designed Kalman filter. In this paper, we evaluate different multi-sensor fusion systems from the perspective of observability. To adjust and optimize the filter performance before filtering, in this paper, we derive the expression form of estimation error covariance of three different fusion methods and discussed both observable degree of fusion center and local filter of fusion step. Based on the ODAEPM, we obtained their discriminant matrix of observable degree and the relationship among different fusion methods is given by mathematical proof. To confirm mathematical conclusion, the simulation analysis is done for multi-sensor CV model. The result demonstrates our theory and verifies the advantage of information fusion system.

Nash Bargaining Game-Theoretic Framework for Power Control in Distributed Multiple-Radar Architecture Underlying Wireless Communication System.

This paper presents a novel Nash bargaining solution (NBS)-based cooperative game-theoretic framework for power control in a distributed multiple-radar architecture underlying a wireless communication system. Our primary objective is to minimize the total power consumption of the distributed multiple-radar system (DMRS) with the protection of wireless communication user's transmission, while guaranteeing each radar's target detection requirement. A unified cooperative game-theoretic framework is proposed for the optimization problem, where interference power constraints (IPCs) are imposed to protect the communication user's transmission, and a minimum signal-to-interference-plus-noise ratio (SINR) requirement is employed to provide reliable target detection for each radar. The existence, uniqueness and fairness of the NBS to this cooperative game are proven. An iterative Nash bargaining power control algorithm with low computational complexity and fast convergence is developed and is shown to converge to a Pareto-optimal equilibrium for the cooperative game model. Numerical simulations and analyses are further presented to highlight the advantages and testify to the efficiency of our proposed cooperative game algorithm. It is demonstrated that the distributed algorithm is effective for power control and could protect the communication system with limited implementation overhead.

Functionalization of C-H Bonds by Photoredox Catalysis.

Visible-light photoredox catalysis has been successfully used in the functionalization of inert C-H bonds including C(sp2 )-H bonds of arenes and C(sp3 )-H bonds of aliphatic compounds over the past decade. These transformations are typically promoted by the process of single-electron-transfer (SET) between substrates and photo-excited photocatalyst upon visible light irradiation (household bulbs or LEDs). Compared with other synthetic strategies, such as the transition-metal catalysis and traditional radical reactions, visible-light photoredox approach has distinct advantages in terms of operational simplicity and practicability. Versatile direct functionalization of inert C(sp2 )-H and C(sp3 )-H bonds including alkylation, trifluoromethylation, arylation and amidation, has been achieved using this practical strategy.