Dual-Path Vision Transformer用于急性缺血性脑卒中辅助诊断

急性缺血性脑卒中是由于脑组织血液供应障碍导致的脑功能障碍,数字减影脑血管造影(DSA)是诊断脑血管疾病的金标准。基于患者的正面和侧面DSA图像,对急性缺血性脑卒中的治疗效果进行分级评估,构建基于Vision Transformer的双路径图像分类智能模型DPVF。为了提高辅助诊断速度,基于EdgeViT的轻量化设计思想进行了模型的构建;为了使模型保持轻量化的同时具有较高的精度,提出空间-通道自注意力模块,促进Transformer模型捕获更全面的特征信息,提高模型的表达能力;此外,对于DPVF的两分支的特征融合,构建交叉注意力模块对两分支输出进行交叉融合,促使模型提取更丰富的特征,从而提高模型表现。实验结果显示DPVF在测试集上的准确率达98.5%,满足实际需求。

Mechanical properties of fine-grained dual phase low-carbon steels based on dynamic transformation

The fine grained dual phase （FG-DP） steel with ferrite grains of 2-4.5 μm and martensite islands smaller than 3 μm was obtained through the mechanism of deformation-enhanced ferrite transformation （DEFT）. Mechanical properties of the steel were tested at room temperature. The results indicated that with a similar volume fraction of martensite （about 20vol%）,FG-DP steel exhibited a superior combination of higher strength and more rapid strain hardening at low strains compared with the coarse-grained dual phase （CG-DP） steel obtained by critical annealing. The combination of higher strength,large elongation,and more rapid strain hardening of FG-DP steel can be attributed to the fine ferrite grain and finely dispersed martensite islands. In addition,the uniformly distributed martensite islands in FG-DP steel have smaller interspacing compared with that of CG-DP steel. So,at the initial plastic deformation stage,the plastic deformation of ferrite was restrained and more pronounced load was transferred from ferrite to martensite. The plastic deformation of martensite in FG-DP steel started earlier.

DUAL NUMBER TRANSFORMATION AND ITS APPLICATION TO ROBOT WITH REDUNDANCY

This paper presents the application of dual-number matrix to the formulation of Jacobian equations of robot with redundancy, the analytical technique that is based on the dual-number matrices, a 3 × 3 matrix with dualnumber elements, and the dual-number transformation method. Dual-number matrices make possible a concise representation of joint parameters. In particular, the method can effectively be used for direct determination of Jacobian matrices. It is shown that the proposed procedure contributes a simplified approach to the formulation of robotic kinematics, dynamics and control system modelling.

The Dual Functionals for the Generalized Ball Basis of Wang-Said Type and Basis Transformation Formulas

The generalized Ball curves of Wang-Said type with a position parameter L not only unify the Wang-Ball curves and the Said-Ball curves, but also include several useful intermediate curves. This paper presents the dual functionals for the generalized Ball basis of Wang-Said type. The relevant basis transformation formulae are also worked out.

Green Transformation of Chinese Oil and Gas Enterprises in Context of the “Dual Carbon” Goals

2021 was the first year of the“14th Five-year Plan”and the year in which the“30·60 Goals”were officially implemented.In the first year of the“Dual Carbon Goals”,domestic oil and gas enterprises vigorously expanded their new energy businesses while actively developing their main businesses,bravely shouldered their responsibility for supply guarantee and comprehensively accelerated their green transformation,creating typical experience and practices.

CHARACTERISTICS OF MARTENSITIC TRANSFORMATION IN DUAL-PHASE STEELS

A theorectical expression for the driving force and M_(?) point of martensitic transformation has been proposed.The M_(?) values using this expression are in good agreement with that obtained experimentally.It was found that the values of M_(?) and M_(?) are not only related to the carbon content in martensite,but also to the volume fraction of ferrite.

Megacity pathways in China under the dual carbon goal:The case of Shanghai

The pathways to achieving carbon neutrality at the city level are diverse due to varying energy supply and demand conditions.Shanghai faces obstacles such as limited land resources,high costs of renewable energy technologies,and instability of renewable energy.These challenges hinder the city’s efforts to achieve carbon peak and carbon neutrality(dual carbon).Therefore,Shanghai must identify and optimize its development path for renewable energy under the dual carbon goal.We employed the Low Emissions Analysis Platform Shanghai(LEAP-SH)model to simulate the impact of policies,such as industrial upgrading,energy efficiency improvement,energy structure optimization,increased technical innovation on energy,and ecological restoration,on the carbon emission pathways from 2022 to 2060 using five different scenarios.Our results indicate that Shanghai has the potential to achieve carbon neutrality in 2059 by promoting carbon reduction,pollution control,and green expansion.Moreover,we determined that the manufacturing industry;power generation industry;and transportation,storage,and mail services are the three major sectors for emission reduction under the dual carbon goal.Furthermore,the capacity and output of coal-fired power plants will be gradually replaced by offshore wind power in the dual carbon pathway.Finally,this study proposes countermeasures and suggestions for Shanghai to attain the dual carbon goal and high-quality development.

DCFNet:An Effective Dual-Branch Cross-Attention Fusion Network for Medical Image Segmentation

Automatic segmentation of medical images provides a reliable scientific basis for disease diagnosis and analysis.Notably,most existing methods that combine the strengths of convolutional neural networks(CNNs)and Transformers have made significant progress.However,there are some limitations in the current integration of CNN and Transformer technology in two key aspects.Firstly,most methods either overlook or fail to fully incorporate the complementary nature between local and global features.Secondly,the significance of integrating the multiscale encoder features from the dual-branch network to enhance the decoding features is often disregarded in methods that combine CNN and Transformer.To address this issue,we present a groundbreaking dual-branch cross-attention fusion network(DCFNet),which efficiently combines the power of Swin Transformer and CNN to generate complementary global and local features.We then designed the Feature Cross-Fusion(FCF)module to efficiently fuse local and global features.In the FCF,the utilization of the Channel-wise Cross-fusion Transformer(CCT)serves the purpose of aggregatingmulti-scale features,and the Feature FusionModule(FFM)is employed to effectively aggregate dual-branch prominent feature regions from the spatial perspective.Furthermore,within the decoding phase of the dual-branch network,our proposed Channel Attention Block(CAB)aims to emphasize the significance of the channel features between the up-sampled features and the features generated by the FCFmodule to enhance the details of the decoding.Experimental results demonstrate that DCFNet exhibits enhanced accuracy in segmentation performance.Compared to other state-of-the-art(SOTA)methods,our segmentation framework exhibits a superior level of competitiveness.DCFNet’s accurate segmentation of medical images can greatly assist medical professionals in making crucial diagnoses of lesion areas in advance.

基于双流特征融合的肺结节图像分割网络

为了提高对肺结节的准确分割,文中提出一种CSF-UNet的双骨干网络特征提取方法。使用两种不同侧重的骨干网络并行提取图像特征,通过利用ConvNeXt网络提取局部特征,并结合Swin Transformer网络提取全局特征来提升模型的特征提取能力。提出了一种自适应大核融合模块,有效地融合两种不同规格的特征,通过串联两个大核卷积获得更大的感受野和动态选择机制来突出重要的空间区域。在SPPF中融合了ECA通道注意力和密集链接,提出了ESPP模块以进一步挖掘双骨干网络提取的高级语义信息,使网络更加关注重要的特征通道。在LIDC数据集上的实验结果表明,提出的模型在3个指标上优于基本UNet模型以及最近几年其他研究团队提出的针对该数据集的分割网络。最终,CSF-UNet模型实现了78.1%的IoU、87.71%的DSC、87.19%的敏感度和88.23%的精确度。这些结果表明,该模型在肺结节分割方面表现出良好的性能,对医生进行早期肺结节诊断具有重要的临床意义和应用价值。

结合视觉显著性与Dual-PCNN的红外与可见光图像融合

针对现存的红外与可见光图像融合算法亮度不均、目标不突出、对比度不高、细节丢失等问题,结合非下采样剪切波变换(NSST)具有多尺度、最具稀疏表达的特性,显著性检测具有突出红外目标的优势,双通道脉冲耦合神经网络(Dual-PCNN)具有耦合、脉冲同步激发等优点,提出一种基于NSST结合视觉显著性引导Dual-PCNN的图像融合方法。首先,通过NSST分解红外与可见光图像各方向的高频与低频子带系数;然后,低频子带系数采用基于显著性决策图引导Dual-PCNN融合策略,高频子带系数采用改进的空间频率作为优化Dual-PCNN的激励进行融合;最后,经过NSST逆变换得到融合图像。实验结果表明,融合图像红外目标突出且可见光背景细节丰富。该方法相比于其他融合算法在主观评价与客观评价上都有一定程度的改善。

Adaptive Dual-Threshold Edge Detection Based on Wavelet Transform

In order to solve the problems of local maximum modulus extraction and threshold selection in the edge detection of finite resolution digital images, a new wavelet transform based adaptive dual threshold edge detection algorithm is proposed. The local maximum modulus is extracted by linear interpolation in wavelet domain. With the analysis on histogram, the image is filtered with an adaptive dual threshold method, which effectively detects the contours of small structures as well as the boundaries of large objects. A wavelet domain's propagation function is used to further select weak edges. Experimental results have shown the self adaptivity of the threshold to images having the same kind of histogram, and the efficiency even in noise tampered images.

Quantization-Based Robust Image Watermarking Using the Dual Tree Complex Wavelet Transform

Conventional quantization index modulation (QIM) watermarking uses the fixed quantization step size for the host signal.This scheme is not robust against geometric distortions and may lead to poor fidelity in some areas of content.Thus,we proposed a quantization-based image watermarking in the dual tree complex wavelet domain.We took advantages of the dual tree complex wavelets (perfect reconstruction,approximate shift invariance,and directional selectivity).For the case of watermark detecting,the probability of false alarm and probability of false negative were exploited and verified by simulation.Experimental results demonstrate that the proposed method is robust against JPEG compression,additive white Gaussian noise (AWGN),and some kinds of geometric attacks such as scaling,rotation,etc.

MULTIPLIER OPERATORS AND EXTREMAL FUNCTIONS RELATED TO THE DUAL DUNKL-SONINE OPERATOR

We study the dual Dunkl-Sonine operator tSk,e on Rd and give expression of tSk,t, using Dunkl multiplier operators on Rd, Next, we study the extremal functions fλ, λ〉 0 related to the Dunkl multiplier operators, and more precisely show that {fλ}λ〉0 converges uniformly to tSk,e（f） as λ→0 Certain examples based on Dunkl-heat and Dunkt-Poisson kernels are provided to illustrate the results.

A Dual Tree Complex Discrete Cosine Harmonic Wavelet Transform (ADCHWT) and Its Application to Signal/Image Denoising

A new simple and efficient dual tree analytic wavelet transform based on Discrete Cosine Harmonic Wavelet Transform DCHWT (ADCHWT) has been proposed and is applied for signal and image denoising. The analytic DCHWT has been realized by applying DCHWT to the original signal and its Hilbert transform. The shift invariance and the envelope extraction properties of the ADCHWT have been found to be very effective in denoising speech and image signals, compared to that of DCHWT.

Digital Transformation:Reshaping China’s Development Landscape

The Chinese economy is currently undergoing a digital transformation.New growth drivers are replacing old ones,creating a new development landscape.Countries with strong digital industries will be the first to reap the benefits of digitalization.For China,the transformation to a digital economy is both of inevitability and heterogeneity fueled by dual circulations.Crowded out from more skill-based digitalized sectors,less-skilled labor moves to less digitalized sectors.New capital,industries and technology clusters emerge as new drivers of manufacturing and service sector development.With its large domestic market and industrial competitiveness,China has fostered a new development landscape of“dual circulations”.

Variable spot size optical system for a dual-wavelength laser therapy device

The majority of existing high-power laser therapeutic instruments employ a single wavelength for a single target;thus,they do not meet the requirements for clinical treatment.Therefore,this study designs an optical system for a dual-wavelength high-power laser therapeutic device with a variable spot size.The waist of the short arm of the optical cavity and the G1G2 parameter(G-parameter equivalent cavity method)is calculated using MATLAB software,the spot size and divergence angle on the lens are calculated using an ABCD matrix,and the distance between the treatment spot at different spot sizes and the transformation lens is calculated in order to design the treatment handpiece.Experiments are conducted to analyze the stability at an output power of 532 nm before beam combination and the power loss after beam combination.The results show that the output power stability of the 532-nm beam varies by less than 2%over 150 min,and the loss of both wavelengths is less than 20%,which meets the clinical requirements of the system.The safety performance can meet the requirements of national general standards for medical electrical safety.The proposed dual-wavelength laser therapy instrument has both visible wave and near-infrared wave characteristics;thus,it can accurately target both superficial vessels and vessels with a larger diameter and deeper position.This therapeutic device has the advantages of simple operation,stable and reliable laser output,high security and strong anti-interference ability,and meets the comprehensive clinical treat-ment demands of vascular diseases.

DISTANCE MEASURING MODELING AND ERROR ANALYSIS OF DUALCCD VISION SYSTEM SIMULATING HUMAN EYES AND NECK

A dual-CCD simulating human eyes and neck (DSHEN) vision system is putforward. Its structure and principle are introduced. The DSREN vision system can perform somemovements simulating human eyes and neck by means of four rotating joints, and realize preciseobject recognizing and distance measuring in all orientations. The mathematic model of the DSHENvision system is built, and its movement equation is solved. The coordinate error and measureprecision affected by the movement parameters are analyzed by means of intersection measuringmethod. So a theoretic foundation for further research on automatic object recognizing and precisetarget tracking is provided.

Effect of controlled rolling on the martensitic hardenability of dual phase steel

The C-Mn and C-Mn-Nb steels were thermo-mechanically processed to develop dual phase steel and to study the effect of controlled rolling on the martensitic hardenability of austenite. The steel specimens were intercritically annealed at 790℃, rolled at that temperature to the reductions of 10%, 23%, and 47% and immediately cooled at different rates. Quantitative metallography was used to construct the microstructure map, which illustrated that increasing deformation progressively reduced the proportion of new ferrite formed at all cooling rates and increased the amount of martensite at fast and intermediate rates. The martensitic hardenability of austenite remaining after all the rolling reductions was plotted as a function of cooling rates. It was observed that for the austenite-martensite conversion efficiencies greater than about 25%, controlled rolling increased the martensitic hardenability of austenite.

MATHEMATICAL PROBLEMS IN THE INTEGRAL-TRANSFORMATION METHOD OF DYNAMIC CRACK

In the investigation on fracture mechanics,the potential function was introduced, and the moving differential equation was constructed. By making Laplace and Fourier transformation as well as sine and cosine transformation to moving differential equations and various responses, the dual equation which is constructed from boundary conditions lastly was solved. This method of investigating dynamic crack has become a more systematic one that is used widely. Some problems are encountered when the dynamic crack is studied. After the large investigation on the problems, it is discovered that during the process of mathematic derivation, the method is short of precision, and the derived results in this method are accidental and have no credibility.A model for example is taken to explain the problems existing in initial deriving process of the integral_transformation method of dynamic crack.

EXACT SOLUTION AND ITS BEHAVIOR CHARACTERISTIC OF NONLINEAR DUAL-POROSITY MODEL

A nonlinear dual-porosity model considering a quadratic gradient term is presented. Assuming the pressure difference between matrix and fractures as a primary unknown, to avoid solving the simultaneous system of equations, decoupling of fluid pressures in the blocks from the fractures was furnished with a quasi-steady-state flow in the blocks. Analytical solutions were obtained in a radial flow domain using generalized Hankel transform. The real value cannot be gotten because the analytical solutions were infinite series. The real pressure value was obtained by numerical solving the eigenvalue problem. The change law of pressure was studied while the nonlinear parameters and dual-porosity parameters changed, and the plots of typical curves are given. All these result can be applied in well test analysis.