Attaining a highly efficient and inexpensive electrocatalyst is significant for the hydrogen evolution reaction(HER)but still challenging nowadays.The transition-metal phosphides(TMPs)catalysts with platinum-like elec...Attaining a highly efficient and inexpensive electrocatalyst is significant for the hydrogen evolution reaction(HER)but still challenging nowadays.The transition-metal phosphides(TMPs)catalysts with platinum-like electronic structures are a potential candidate for the HER,but those are prone to be strongly bound with hydrogen intermediates(H∗),resulting in sluggish HER kinetics.Herein we report a unique hybrid structure of CoP anchored on graphene nanoscrolls@carbon nano tubes(CNTs)scaffold(Ni M@C-CoP)encapsulating various Ni M(M=Zn,Mo,Ni,Co)bimetal nanoalloy via chemical vapor deposi-tion(CVD)growth of CNT on graphene nanoscrolls followed by the impregnation of cobalt precursors and phosphorization for efficiently electrocatalytic hydrogen evolution.CoP nanoparticles mainly scattered at the tip of CNT branches which exhibited the analogical“Three-layer core-shell”structures.Experiments and density functional theory(DFT)calculations consistently disclose that the encapsulated various NiMs can offer different numbers of electrons to weaken the interactions of outmost CoP with H∗and push the downshift of the d-band center to different degrees as well as stabilize the outmost CoP nanopar-ticles to gain catalytic stability via the electron traversing effect.The electrocatalytic HER activity can be maximumly enhanced with low overpotentials of 78 mV(alkaline)and 89 mV(acidic)at a current density of 10 mA/cm^(2) and sustained at least 24 h especially for NiZn@C-CoP catalyst.This novel system is distinct from conventional three-layer heterostructure,providing a specially thought of d-band center control engineering strategy for the design of heterogeneous catalysts and expanding to other electrocat-alysts,energy storage,sensing,and other applications.展开更多
Co-free Li-rich layered oxides(LLOs)are emerging as promising cathode materials for Li-ion batteries due to their low cost and high capacity.However,they commonly face severe structural instability and poor electroche...Co-free Li-rich layered oxides(LLOs)are emerging as promising cathode materials for Li-ion batteries due to their low cost and high capacity.However,they commonly face severe structural instability and poor electrochemical activity,leading to diminished capacity and voltage performance.Herein,we introduce a Co-free LLO,Li_(1.167)Ni_(0.222)Mn_(0.611)O_(2)(Cf-L1),which features a cooperative structure of Li/Ni mixing and stacking faults.This structure regulates the crystal and electronic structures,resulting in a higher discharge capacity of 300.6 mA h g^(-1)and enhanced rate capability compared to the typical Co-free LLO,Li_(1.2)Ni_(0.2)Mn_(0.6)O_(2)(Cf-Ls).Density functional theory(DFT)indicates that Li/Ni mixing in LLOs leads to increased Li-O-Li configurations and higher anionic redox activities,while stacking faults further optimize the electronic interactions of transition metal(TM)3d and non-bonding O 2p orbitals.Moreover,stacking faults accommodate lattice strain,improving electrochemical reversibility during charge/discharge cycles,as demonstrated by the in situ XRD of Cf-L1 showing less lattice evolution than Cf-Ls.This study offers a structured approach to developing Co-free LLOs with enhanced capacity,voltage,rate capability,and cyclability,significantly impacting the advancement of the next-generation Li-ion batteries.展开更多
Hydrogen is a clean and flexible energy carrier that has the promising to satisfy urgent demands of the energy crisis and environmental protection.Electrochemical hydrogen evolution reaction(HER),a critical half-react...Hydrogen is a clean and flexible energy carrier that has the promising to satisfy urgent demands of the energy crisis and environmental protection.Electrochemical hydrogen evolution reaction(HER),a critical half-reaction in water splitting,is one of the greenest and most common methods to obtain high-purity hydrogen.Designing preeminent activity and stability electrocatalysts for hydrogen precipitation reac-tion(HER)to reduce energy consumption is of great essential.3D carbon-based materials have attracted widespread concern as the potential scaffolds of highly active and durable electrocatalysts for HER.To boost the HER activity and prolong the lifespan of electrocatalysts,multifarious 3D carbon architectures make an appearance to be engineered for accelerating electronic/mass transfer and maximizing the expo-sure of active sites.Herein,we designed and fabricated high-performance electrocatalysts based on a spe-cial caterpillar-like 3D graphene nanoscrolls@CNTs(GNS@CNTs)scaffold decorated with Co-doped MoSe_(2)nanosheets for HER.In the caterpillar-like hierarchical structure,CNTs were seamlessly co-bonded and dilated the interlayer and outer spacing of GNS through CVD growth technology,and nickel nanoparticles were covered by the CNTs tips.Taking advantage of the plentiful hierarchical pore,larger specific surface area,and higher chemical stability of the caterpillar-like structure,the catalysts exhibited enhanced elec-trocatalytic properties than some existing data reported.Density functional theory calculations showed that the encapsulated nickel nanoparticle could tune the electronic structure of the outer anchored Co-doped MoSe_(2)and optimize itsG of H∗adsorption by electron traversing effect and doping effect.These indicate that caterpillar-like GNS@CNT is an ideal scaffold f or anchoring actives substance and is suit-able for high-efficient HER.This study provides new insights for designing hierarchical carbon composite nanostructures for catalysts,sensors,energy materials,and other applications.展开更多
Based on the star formation histories of galaxies in halos with different masses, we develop an empirical model to grow galaxies in dark matter halos. This model has very few ingredients, any of which can be associate...Based on the star formation histories of galaxies in halos with different masses, we develop an empirical model to grow galaxies in dark matter halos. This model has very few ingredients, any of which can be associated with observational data and thus be efficiently assessed. By applying this model to a very high resolution cosmological N-body simulation, we predict a number of galaxy properties that are a very good match to relevant observational data. Namely, for both centrals and satellites, the galaxy stellar mass functions up to redshift z=4 and the conditional stellar mass functions in the local universe are in good agreement with observations. In addition, the two point correlation function is well predicted in the different stellar mass ranges explored by our model. Furthermore, after applying stellar population synthesis models to our stellar composition as a function of redshift, we find that the luminosity functions in the 0.1 u,0.19, 0.1r, 0.1i and 0.1z bands agree quite well with the SDSS observational results down to an absolute magnitude at about -17.0. The SDSS conditional luminosity function itself is predicted well. Finally, the cold gas is derived from the star formation rate to predict the HI gas mass within each mock galaxy. We find a remarkably good match to observed HI-to-stellar mass ratios. These features ensure that such galaxy/gas catalogs can be used to generate reliable mock redshift surveys.展开更多
Medical image registration is vital for disease diagnosis and treatment with its ability to merge diverse informa-tion of images,which may be captured under different times,angles,or modalities.Although several survey...Medical image registration is vital for disease diagnosis and treatment with its ability to merge diverse informa-tion of images,which may be captured under different times,angles,or modalities.Although several surveys have reviewed the development of medical image registration,they have not systematically summarized the existing med-ical image registration methods.To this end,a comprehensive review of these methods is provided from traditional and deep-learning-based perspectives,aiming to help audiences quickly understand the development of medical image registration.In particular,we review recent advances in retinal image registration,which has not attracted much attention.In addition,current challenges in retinal image registration are discussed and insights and prospects for future research provided.展开更多
Deformable retinal image registration is crucial in clinical diagnosis and longitudinal studies of retinal diseases.Most existing deep deformable retinal image registration methods focus on fully convolutional network...Deformable retinal image registration is crucial in clinical diagnosis and longitudinal studies of retinal diseases.Most existing deep deformable retinal image registration methods focus on fully convolutional network(FCN)architecture design,which fails to model long-range dependencies among pixels-a significant factor in deformable retinal image registration.Transformers based on the self-attention mechanism,can capture global context dependencies,complementing local convolution.However,multi-scale spatial feature fusion and pixel-wise position selection are also crucial for the deformable retinal image registration,are often ignored by both FCNs and transformers.To fully leverage the merits of FCNs,multi-scale spatial attention and transformers,we propose a hierarchical hybrid architecture,reparameterized multi-scale transformer(RMFormer),for deformable retinal image registration.In RMFormer,we specifically develop a reparameterized multi-scale spatial attention to adaptively fuse multi-scale spatial features,with the assistance of the reparameterizing technique,thereby highlighting informative pixel-wise positions in a lightweight manner.The experimental results on two publicly available datasets demonstrate the superiority of our RMFormer over state-of-the-art methods and show that it is data-efficient in a limited medical image regime.Additionally,we are the first to provide a visualization analysis to explain how our proposed method affects the deformable retinal image registration process.The source code of our work is available at https://github.com/Tloops/RMFormer.展开更多
Seagrass meadows are generally diverse in China and have become important ecosystem with essential functions.However,the seagrass distribution across the seawaters of China has not been evaluated,and the magnitude and...Seagrass meadows are generally diverse in China and have become important ecosystem with essential functions.However,the seagrass distribution across the seawaters of China has not been evaluated,and the magnitude and direction of changes in seagrass meadows remain unclear.This study aimed to provide a nationwide seagrass distribution map and explore the dynamic changes in seagrass population under global climate change.Simulation studies were performed using the modeling software MaxEnt with 58961 occurrence records and 27 marine environmental variables to predict the potential distribution of seagrasses and calculate the area.Seven environmental variables were excluded from the modeling processes based on a correlation analysis to ensure predicted suitability.The predicted area was 790.09 km^(2),which is much larger than the known seagrass distribution in China(87.65 km^(2)).By 2100,the suitable habitat of seagrass will shift northwest and increase to 923.62 km2.Models of the sum of the individual family under-pre-dicted the national distribution of seagrasses and consistently showed a downward trend in the future.Out of all environmental vari-ables,physical parameters(e.g.,depth,land distance,and sea surface temperature)contributed the most in predicting seagrass distri-butions,and nutrients(e.g.,nitrate,phosphate)ranked among the key influential predictors for habitat suitability in our study area.This study is the first effort to fill a gap in understanding the distribution of seagrasses in China.Further studies using modeling and biological/ecological approaches are warranted.展开更多
Nowadays,high-performance computing(HPC)clusters are increasingly popular.Large volumes of job logs recording many years of operation traces have been accumulated.In the same time,the HPC cloud makes it possible to ac...Nowadays,high-performance computing(HPC)clusters are increasingly popular.Large volumes of job logs recording many years of operation traces have been accumulated.In the same time,the HPC cloud makes it possible to access HPC services remotely.For executing applications,both HPC end-users and cloud users need to request specific resources for different workloads by themselves.As users are usually not familiar with the hardware details and software layers,as well as the performance behavior of the underlying HPC systems.It is hard for them to select optimal resource configurations in terms of performance,cost,and energy efficiency.Hence,how to provide on-demand services with intelligent resource allocation is a critical issue in the HPC community.Prediction of job characteristics plays a key role for intelligent resource allocation.This paper presents a survey of the existing work and future directions for prediction of job characteristics for intelligent resource allocation in HPC systems.We first review the existing techniques in obtaining performance and energy consumption data of jobs.Then we survey the techniques for single-objective oriented predictions on runtime,queue time,power and energy consumption,cost and optimal resource configuration for input jobs,as well as multi-objective oriented predictions.We conclude after discussing future trends,research challenges and possible solutions towards intelligent resource allocation in HPC systems.展开更多
Gastric cancer(GC)is one of the most common cancers and ranks the third in cancer mortality all over the world.The goal of this study was to identify potential hub-genes,highlighting their functions,signaling pathways...Gastric cancer(GC)is one of the most common cancers and ranks the third in cancer mortality all over the world.The goal of this study was to identify potential hub-genes,highlighting their functions,signaling pathways,and candidate drugs for the treatment of GC patients.We used publicly available next generation sequencing(NGS)data to identify differentially expressed(DE)genes.The top DE genes were mapped to STRING database to construct the protein-protein interaction(PPI)network and top hub genes were selected for further analysis.We found a total of 1555 DE genes with 870 upregulated and 685 downregulated genes in GC.We selected the top 400(200 upregulated and 200 downregulated)genes to construct a PPI network and extracted the top 15 hub genes.The gene ontology(GO)term and kyoto encyclopedia of genes and genomes(KEGG)pathway enrichment analyses of the 15 hub genes exposed some important functions and signaling pathways that were significantly associated with GC patients.The survival analysis of the hub genes disclosed that the lower expressions of the three hub genes CDH2,COL4A1,and COL5A2 were associated with better survival of GC patients.These three genes might be the candidate biomarkers for the diagnosis and treatment of GC.Then,we considered 3 key proteins(genomic biomarkers)(COL4A1,CDH2,and CO5A2)as the drug target proteins(receptors),performed their docking analysis with the 102 meta-drug agents,and found Everolimus,Docetaxel,Lanreotide,Venetoclax,Temsirolimus,and Nilotinib as the top ranked 6 candidate drugs with respect to our proposed target proteins for the treatment against GC patients.Therefore,the proposed drugs might play vital role for the treatment against GC patients.展开更多
Recently,analyzing big data on the move is booming.It requires that the hardware resource should be low volume,low power,light in weight,high-performance,and highly scalable whereas the management software should be f...Recently,analyzing big data on the move is booming.It requires that the hardware resource should be low volume,low power,light in weight,high-performance,and highly scalable whereas the management software should be flexible and consume little hardware resource.To meet these requirements,we present a system named SOCA-DOM that encompasses a mobile system-on-chip array architecture and a two-tier“software-defined”resource manager named Chameleon.First,we design an Ethernet communication board to support an array of mobile system-on-chips.Second,we propose a two-tier software architecture for Chameleon to make it flexible.Third,we devise data,configuration,and control planes for Chameleon to make it“software-defined”and in turn consume hardware resources on demand.Fourth,we design an accurate synthetic metric that represents the computational power of a computing node.We employ 12 Apache Spark benchmarks to evaluate SOCA-DOM.Surprisingly,SOCA-DOM consumes up to 9.4x less CPU resources and 13.5x less memory than Mesos which is an existing resource manager.In addition,we show that a 16-node SOCA-DOM consumes up to 4x less energy than two standard Xeon servers.Based on the results,we conclude that an array architecture with fine-grained hardware resources and a software-defined resource manager works well for analyzing big data on the move.展开更多
Autism Spectrum Disorder(ASD)is a highly disabling mental disease that brings significant impairments of social interaction ability to the patients,making early screening and intervention of ASD critical.With the deve...Autism Spectrum Disorder(ASD)is a highly disabling mental disease that brings significant impairments of social interaction ability to the patients,making early screening and intervention of ASD critical.With the development of the machine learning and neuroimaging technology,extensive research has been conducted on machine classification of ASD based on structural Magnetic Resonance Imaging(s-MRI).However,most studies involve with datasets where participants'age are above 5 and lack interpretability.In this paper,we propose a machine learning method for ASD classification in children with age range from 0.92 to 4.83 years,based on s-MRI features extracted using Contrastive Variational AutoEncoder(CVAE).78 s-MRIs,collected from Shenzhen Children's Hospital,are used for training CVAE,which consists of both ASD-specific feature channel and common-shared feature channel.The ASD participants represented by ASD-specific features can be easily discriminated from Typical Control(TC)participants represented by the common-shared features.In case of degraded predictive accuracy when data size is extremely small,a transfer learning strategy is proposed here as a potential solution.Finally,we conduct neuroanatomical interpretation based on the correlation between s-MRI features extracted from CVAE and surface area of different cortical regions,which discloses potential biomarkers that could help target treatments of ASD in the future.展开更多
Improving performance of deep learning models and reducing their training times are ongoing challenges in deep neural networks.There are several approaches proposed to address these challenges,one of which is to incre...Improving performance of deep learning models and reducing their training times are ongoing challenges in deep neural networks.There are several approaches proposed to address these challenges,one of which is to increase the depth of the neural networks.Such deeper networks not only increase training times,but also suffer from vanishing gradients problem while training.In this work,we propose gradient amplification approach for training deep learning models to prevent vanishing gradients and also develop a training strategy to enable or disable gradient amplification method across several epochs with different learning rates.We perform experiments on VGG-19 and Resnet models(Resnet-18 and Resnet-34),and study the impact of amplification parameters on these models in detail.Our proposed approach improves performance of these deep learning models even at higher learning rates,thereby allowing these models to achieve higher performance with reduced training time.展开更多
The main goal of this paper is to develop the coupled double-distributionfunction(DDF)lattice Boltzmann method(LBM)for simulation of subsonic and transonic turbulent flows.In the present study,we adopt the second-orde...The main goal of this paper is to develop the coupled double-distributionfunction(DDF)lattice Boltzmann method(LBM)for simulation of subsonic and transonic turbulent flows.In the present study,we adopt the second-order implicit-explicit(IMEX)Runge-Kutta schemes for time discretization and the Non-Oscillatory and NonFree-Parameters Dissipative(NND)finite difference scheme for space discretization.The Sutherland’s law is used for expressing the viscosity of the fluid due to considerable temperature change.Also,the Spalart-Allmaras(SA)turbulence model is incorporated in order for the turbulent flow effect to be pronounced.Numerical experiments are performed on different turbulent compressible flows around a NACA0012 airfoil with body-fitted grid.Our numerical results are found to be in good agreement with experiment data and/or other numerical solutions,demonstrating the applicability of the method presented in this study to simulations of both subsonic and transonic turbulent flows.展开更多
基金This work was supported by the Science and Technology Pro-gram of Shaanxi Province(No.2019GY-200).Shengwu Guo and Wei Wang contributed to the material TEM and SEM characterizations in this work.
文摘Attaining a highly efficient and inexpensive electrocatalyst is significant for the hydrogen evolution reaction(HER)but still challenging nowadays.The transition-metal phosphides(TMPs)catalysts with platinum-like electronic structures are a potential candidate for the HER,but those are prone to be strongly bound with hydrogen intermediates(H∗),resulting in sluggish HER kinetics.Herein we report a unique hybrid structure of CoP anchored on graphene nanoscrolls@carbon nano tubes(CNTs)scaffold(Ni M@C-CoP)encapsulating various Ni M(M=Zn,Mo,Ni,Co)bimetal nanoalloy via chemical vapor deposi-tion(CVD)growth of CNT on graphene nanoscrolls followed by the impregnation of cobalt precursors and phosphorization for efficiently electrocatalytic hydrogen evolution.CoP nanoparticles mainly scattered at the tip of CNT branches which exhibited the analogical“Three-layer core-shell”structures.Experiments and density functional theory(DFT)calculations consistently disclose that the encapsulated various NiMs can offer different numbers of electrons to weaken the interactions of outmost CoP with H∗and push the downshift of the d-band center to different degrees as well as stabilize the outmost CoP nanopar-ticles to gain catalytic stability via the electron traversing effect.The electrocatalytic HER activity can be maximumly enhanced with low overpotentials of 78 mV(alkaline)and 89 mV(acidic)at a current density of 10 mA/cm^(2) and sustained at least 24 h especially for NiZn@C-CoP catalyst.This novel system is distinct from conventional three-layer heterostructure,providing a specially thought of d-band center control engineering strategy for the design of heterogeneous catalysts and expanding to other electrocat-alysts,energy storage,sensing,and other applications.
基金financially supported by the National Natural Science Foundation of China(52202046,51602246,and 51801144)the Natural Science Foundation of Shanxi Provincial(2021JQ-034)。
文摘Co-free Li-rich layered oxides(LLOs)are emerging as promising cathode materials for Li-ion batteries due to their low cost and high capacity.However,they commonly face severe structural instability and poor electrochemical activity,leading to diminished capacity and voltage performance.Herein,we introduce a Co-free LLO,Li_(1.167)Ni_(0.222)Mn_(0.611)O_(2)(Cf-L1),which features a cooperative structure of Li/Ni mixing and stacking faults.This structure regulates the crystal and electronic structures,resulting in a higher discharge capacity of 300.6 mA h g^(-1)and enhanced rate capability compared to the typical Co-free LLO,Li_(1.2)Ni_(0.2)Mn_(0.6)O_(2)(Cf-Ls).Density functional theory(DFT)indicates that Li/Ni mixing in LLOs leads to increased Li-O-Li configurations and higher anionic redox activities,while stacking faults further optimize the electronic interactions of transition metal(TM)3d and non-bonding O 2p orbitals.Moreover,stacking faults accommodate lattice strain,improving electrochemical reversibility during charge/discharge cycles,as demonstrated by the in situ XRD of Cf-L1 showing less lattice evolution than Cf-Ls.This study offers a structured approach to developing Co-free LLOs with enhanced capacity,voltage,rate capability,and cyclability,significantly impacting the advancement of the next-generation Li-ion batteries.
基金This work was financially supported by the Science and Tech-nology Program of Shaanxi Province(No.2019GY-200).Shengwu Guo contributed to the material TEM characterization in this work.
文摘Hydrogen is a clean and flexible energy carrier that has the promising to satisfy urgent demands of the energy crisis and environmental protection.Electrochemical hydrogen evolution reaction(HER),a critical half-reaction in water splitting,is one of the greenest and most common methods to obtain high-purity hydrogen.Designing preeminent activity and stability electrocatalysts for hydrogen precipitation reac-tion(HER)to reduce energy consumption is of great essential.3D carbon-based materials have attracted widespread concern as the potential scaffolds of highly active and durable electrocatalysts for HER.To boost the HER activity and prolong the lifespan of electrocatalysts,multifarious 3D carbon architectures make an appearance to be engineered for accelerating electronic/mass transfer and maximizing the expo-sure of active sites.Herein,we designed and fabricated high-performance electrocatalysts based on a spe-cial caterpillar-like 3D graphene nanoscrolls@CNTs(GNS@CNTs)scaffold decorated with Co-doped MoSe_(2)nanosheets for HER.In the caterpillar-like hierarchical structure,CNTs were seamlessly co-bonded and dilated the interlayer and outer spacing of GNS through CVD growth technology,and nickel nanoparticles were covered by the CNTs tips.Taking advantage of the plentiful hierarchical pore,larger specific surface area,and higher chemical stability of the caterpillar-like structure,the catalysts exhibited enhanced elec-trocatalytic properties than some existing data reported.Density functional theory calculations showed that the encapsulated nickel nanoparticle could tune the electronic structure of the outer anchored Co-doped MoSe_(2)and optimize itsG of H∗adsorption by electron traversing effect and doping effect.These indicate that caterpillar-like GNS@CNT is an ideal scaffold f or anchoring actives substance and is suit-able for high-efficient HER.This study provides new insights for designing hierarchical carbon composite nanostructures for catalysts,sensors,energy materials,and other applications.
基金supported by the 973 Program(No.2015CB857002)the National Natural Science Foundation of China(Grant Nos.11203054,11128306,11121062,11233005,11073017and 11421303)+2 种基金NCET-11-0879,the Strategic Priority Research Program“The Emergence of Cosmological Structures”of the Chinese Academy of Sciences,Grant No.XDB09000000the Shanghai Committee of Science and Technology,China(Grant No.12ZR1452800)supported by the High Performance Computing Resource in the Core Facility for Advanced Research Computing at Shanghai Astronomical Observatory
文摘Based on the star formation histories of galaxies in halos with different masses, we develop an empirical model to grow galaxies in dark matter halos. This model has very few ingredients, any of which can be associated with observational data and thus be efficiently assessed. By applying this model to a very high resolution cosmological N-body simulation, we predict a number of galaxy properties that are a very good match to relevant observational data. Namely, for both centrals and satellites, the galaxy stellar mass functions up to redshift z=4 and the conditional stellar mass functions in the local universe are in good agreement with observations. In addition, the two point correlation function is well predicted in the different stellar mass ranges explored by our model. Furthermore, after applying stellar population synthesis models to our stellar composition as a function of redshift, we find that the luminosity functions in the 0.1 u,0.19, 0.1r, 0.1i and 0.1z bands agree quite well with the SDSS observational results down to an absolute magnitude at about -17.0. The SDSS conditional luminosity function itself is predicted well. Finally, the cold gas is derived from the star formation rate to predict the HI gas mass within each mock galaxy. We find a remarkably good match to observed HI-to-stellar mass ratios. These features ensure that such galaxy/gas catalogs can be used to generate reliable mock redshift surveys.
基金supported in part by General Program of National Natural Science Foundation of China,Nos.82102189 and 82272086Guangdong Provincial Department of Education,No.SJZLGC202202.
文摘Medical image registration is vital for disease diagnosis and treatment with its ability to merge diverse informa-tion of images,which may be captured under different times,angles,or modalities.Although several surveys have reviewed the development of medical image registration,they have not systematically summarized the existing med-ical image registration methods.To this end,a comprehensive review of these methods is provided from traditional and deep-learning-based perspectives,aiming to help audiences quickly understand the development of medical image registration.In particular,we review recent advances in retinal image registration,which has not attracted much attention.In addition,current challenges in retinal image registration are discussed and insights and prospects for future research provided.
基金supported in part by the National Natural Science Foundation of China(No.82272086)the Leading Goose Program of Zhejiang,China(No.2023C03079)the Shenzhen Natural Science Fund,China(No.JCYJ20200109140820699).
文摘Deformable retinal image registration is crucial in clinical diagnosis and longitudinal studies of retinal diseases.Most existing deep deformable retinal image registration methods focus on fully convolutional network(FCN)architecture design,which fails to model long-range dependencies among pixels-a significant factor in deformable retinal image registration.Transformers based on the self-attention mechanism,can capture global context dependencies,complementing local convolution.However,multi-scale spatial feature fusion and pixel-wise position selection are also crucial for the deformable retinal image registration,are often ignored by both FCNs and transformers.To fully leverage the merits of FCNs,multi-scale spatial attention and transformers,we propose a hierarchical hybrid architecture,reparameterized multi-scale transformer(RMFormer),for deformable retinal image registration.In RMFormer,we specifically develop a reparameterized multi-scale spatial attention to adaptively fuse multi-scale spatial features,with the assistance of the reparameterizing technique,thereby highlighting informative pixel-wise positions in a lightweight manner.The experimental results on two publicly available datasets demonstrate the superiority of our RMFormer over state-of-the-art methods and show that it is data-efficient in a limited medical image regime.Additionally,we are the first to provide a visualization analysis to explain how our proposed method affects the deformable retinal image registration process.The source code of our work is available at https://github.com/Tloops/RMFormer.
基金supported by the National Key R&D Program of China(No.2019YFC1408405-02)the National Natural Science Foundation of China(No.6207070555)the Youth Foundation of the Shandong Academy of Sciences(No.2019QN0026).
文摘Seagrass meadows are generally diverse in China and have become important ecosystem with essential functions.However,the seagrass distribution across the seawaters of China has not been evaluated,and the magnitude and direction of changes in seagrass meadows remain unclear.This study aimed to provide a nationwide seagrass distribution map and explore the dynamic changes in seagrass population under global climate change.Simulation studies were performed using the modeling software MaxEnt with 58961 occurrence records and 27 marine environmental variables to predict the potential distribution of seagrasses and calculate the area.Seven environmental variables were excluded from the modeling processes based on a correlation analysis to ensure predicted suitability.The predicted area was 790.09 km^(2),which is much larger than the known seagrass distribution in China(87.65 km^(2)).By 2100,the suitable habitat of seagrass will shift northwest and increase to 923.62 km2.Models of the sum of the individual family under-pre-dicted the national distribution of seagrasses and consistently showed a downward trend in the future.Out of all environmental vari-ables,physical parameters(e.g.,depth,land distance,and sea surface temperature)contributed the most in predicting seagrass distri-butions,and nutrients(e.g.,nitrate,phosphate)ranked among the key influential predictors for habitat suitability in our study area.This study is the first effort to fill a gap in understanding the distribution of seagrasses in China.Further studies using modeling and biological/ecological approaches are warranted.
基金This work was partly supported by the National Key R&D Program of China(2018YFB0204100)the Science&Technology Innovation Project of Shaanxi Province(2019ZDLGY17-02)the Fundamental Research Funds for the Central Universities.
文摘Nowadays,high-performance computing(HPC)clusters are increasingly popular.Large volumes of job logs recording many years of operation traces have been accumulated.In the same time,the HPC cloud makes it possible to access HPC services remotely.For executing applications,both HPC end-users and cloud users need to request specific resources for different workloads by themselves.As users are usually not familiar with the hardware details and software layers,as well as the performance behavior of the underlying HPC systems.It is hard for them to select optimal resource configurations in terms of performance,cost,and energy efficiency.Hence,how to provide on-demand services with intelligent resource allocation is a critical issue in the HPC community.Prediction of job characteristics plays a key role for intelligent resource allocation.This paper presents a survey of the existing work and future directions for prediction of job characteristics for intelligent resource allocation in HPC systems.We first review the existing techniques in obtaining performance and energy consumption data of jobs.Then we survey the techniques for single-objective oriented predictions on runtime,queue time,power and energy consumption,cost and optimal resource configuration for input jobs,as well as multi-objective oriented predictions.We conclude after discussing future trends,research challenges and possible solutions towards intelligent resource allocation in HPC systems.
基金This work was partly supported by the National Key Research and Development Program of China(No.2018YFB0204403)Key Research and Development Project of Guangdong Province(No.2021B0101310002)+4 种基金Strategic Priority CAS Project(No.XDB38050100)National Science Foundation of China(No.U1813203)the Shenzhen Basic Research Fund(Nos.RCYX2020071411473419,KQTD20200820113106007,and JSGG20201102163800001)CAS Key Lab(No.2011DP173015)the Youth Innovation Promotion Association(No.Y2021101).
文摘Gastric cancer(GC)is one of the most common cancers and ranks the third in cancer mortality all over the world.The goal of this study was to identify potential hub-genes,highlighting their functions,signaling pathways,and candidate drugs for the treatment of GC patients.We used publicly available next generation sequencing(NGS)data to identify differentially expressed(DE)genes.The top DE genes were mapped to STRING database to construct the protein-protein interaction(PPI)network and top hub genes were selected for further analysis.We found a total of 1555 DE genes with 870 upregulated and 685 downregulated genes in GC.We selected the top 400(200 upregulated and 200 downregulated)genes to construct a PPI network and extracted the top 15 hub genes.The gene ontology(GO)term and kyoto encyclopedia of genes and genomes(KEGG)pathway enrichment analyses of the 15 hub genes exposed some important functions and signaling pathways that were significantly associated with GC patients.The survival analysis of the hub genes disclosed that the lower expressions of the three hub genes CDH2,COL4A1,and COL5A2 were associated with better survival of GC patients.These three genes might be the candidate biomarkers for the diagnosis and treatment of GC.Then,we considered 3 key proteins(genomic biomarkers)(COL4A1,CDH2,and CO5A2)as the drug target proteins(receptors),performed their docking analysis with the 102 meta-drug agents,and found Everolimus,Docetaxel,Lanreotide,Venetoclax,Temsirolimus,and Nilotinib as the top ranked 6 candidate drugs with respect to our proposed target proteins for the treatment against GC patients.Therefore,the proposed drugs might play vital role for the treatment against GC patients.
基金the Key Research and Development Program of Guangdong Province of China under Grant No.2019B010155003the National Natural Science Foundation of China under Grant Nos.61672511,61702495,and 61802384the Shenzhen Institute of Artificial Intelligence and Robotics for Society,The Chinese University of Hong Kong,Shenzhen,and the Alibaba Innovative Research Project for Large-Scale Graph Pattern Discovery,Analysis,and Query Techniques.
文摘Recently,analyzing big data on the move is booming.It requires that the hardware resource should be low volume,low power,light in weight,high-performance,and highly scalable whereas the management software should be flexible and consume little hardware resource.To meet these requirements,we present a system named SOCA-DOM that encompasses a mobile system-on-chip array architecture and a two-tier“software-defined”resource manager named Chameleon.First,we design an Ethernet communication board to support an array of mobile system-on-chips.Second,we propose a two-tier software architecture for Chameleon to make it flexible.Third,we devise data,configuration,and control planes for Chameleon to make it“software-defined”and in turn consume hardware resources on demand.Fourth,we design an accurate synthetic metric that represents the computational power of a computing node.We employ 12 Apache Spark benchmarks to evaluate SOCA-DOM.Surprisingly,SOCA-DOM consumes up to 9.4x less CPU resources and 13.5x less memory than Mesos which is an existing resource manager.In addition,we show that a 16-node SOCA-DOM consumes up to 4x less energy than two standard Xeon servers.Based on the results,we conclude that an array architecture with fine-grained hardware resources and a software-defined resource manager works well for analyzing big data on the move.
基金supported by the Shenzhen Science and Technology Program(Nos.KQTD20200820113106007 and SGDX20201103095603009)the Strategic Priority Research Program of Chinese Academy of Sciences(No.XDB38050100)+4 种基金the Shenzhen Key Laboratory of Intelligent Bioinformatics(No.ZDSYS20220422103800001)the Shenzhen Basic Research Fund(No.RCYX20200714114734194)the Key Research and Development Project of Guangdong Province(No.2021B0101310002)the National Natural Science Foundation of China(Nos.U22A2041 and 62272449)the Youth Innovation Promotion Association(No.Y2021101).
文摘Autism Spectrum Disorder(ASD)is a highly disabling mental disease that brings significant impairments of social interaction ability to the patients,making early screening and intervention of ASD critical.With the development of the machine learning and neuroimaging technology,extensive research has been conducted on machine classification of ASD based on structural Magnetic Resonance Imaging(s-MRI).However,most studies involve with datasets where participants'age are above 5 and lack interpretability.In this paper,we propose a machine learning method for ASD classification in children with age range from 0.92 to 4.83 years,based on s-MRI features extracted using Contrastive Variational AutoEncoder(CVAE).78 s-MRIs,collected from Shenzhen Children's Hospital,are used for training CVAE,which consists of both ASD-specific feature channel and common-shared feature channel.The ASD participants represented by ASD-specific features can be easily discriminated from Typical Control(TC)participants represented by the common-shared features.In case of degraded predictive accuracy when data size is extremely small,a transfer learning strategy is proposed here as a potential solution.Finally,we conduct neuroanatomical interpretation based on the correlation between s-MRI features extracted from CVAE and surface area of different cortical regions,which discloses potential biomarkers that could help target treatments of ASD in the future.
基金supported in part by an NVIDIA Academic Hardware Grant
文摘Improving performance of deep learning models and reducing their training times are ongoing challenges in deep neural networks.There are several approaches proposed to address these challenges,one of which is to increase the depth of the neural networks.Such deeper networks not only increase training times,but also suffer from vanishing gradients problem while training.In this work,we propose gradient amplification approach for training deep learning models to prevent vanishing gradients and also develop a training strategy to enable or disable gradient amplification method across several epochs with different learning rates.We perform experiments on VGG-19 and Resnet models(Resnet-18 and Resnet-34),and study the impact of amplification parameters on these models in detail.Our proposed approach improves performance of these deep learning models even at higher learning rates,thereby allowing these models to achieve higher performance with reduced training time.
基金financially supported mainly by the Aeronautical Science Fund of China(Grant No.20061453020)The funds from the Foundation for Basic Research of Northwestern Polytechnical University,P.R.Chinafrom the Discovery Grant of the Natural Sciences and Engineering Research Council of Canada(NSERC)were also used to support this research work.
文摘The main goal of this paper is to develop the coupled double-distributionfunction(DDF)lattice Boltzmann method(LBM)for simulation of subsonic and transonic turbulent flows.In the present study,we adopt the second-order implicit-explicit(IMEX)Runge-Kutta schemes for time discretization and the Non-Oscillatory and NonFree-Parameters Dissipative(NND)finite difference scheme for space discretization.The Sutherland’s law is used for expressing the viscosity of the fluid due to considerable temperature change.Also,the Spalart-Allmaras(SA)turbulence model is incorporated in order for the turbulent flow effect to be pronounced.Numerical experiments are performed on different turbulent compressible flows around a NACA0012 airfoil with body-fitted grid.Our numerical results are found to be in good agreement with experiment data and/or other numerical solutions,demonstrating the applicability of the method presented in this study to simulations of both subsonic and transonic turbulent flows.
