To investigate the mechanisms of microwave induced pacemaker cell injuries, Wistar rats and the primary pacemaker cells of newborn Wistar rats were exposed to microwave at average power density of 50 mW/cm2. Slower sp...To investigate the mechanisms of microwave induced pacemaker cell injuries, Wistar rats and the primary pacemaker cells of newborn Wistar rats were exposed to microwave at average power density of 50 mW/cm2. Slower spontaneous beating rate, intercellular Ca2＋ aggregation and cell membrane perforation were detected immediately after the exposure. Moreover, hyperpolarizationactivated cyclic nucleotide-gated cation channel 4 （HCN4） was down-regulated immediately after the exposure and up-regulated at 12 h after the exposure. In the sinoatrial node （SAN） of the rats,展开更多
A portable 4-channel electrochemical noise(EN) test system with high precision was developed.The modular instrument cRIO was used as its core and the signal conditioning module included zero resistance ammeter(ZRA),fl...A portable 4-channel electrochemical noise(EN) test system with high precision was developed.The modular instrument cRIO was used as its core and the signal conditioning module included zero resistance ammeter(ZRA),fly line,screening box and shielded wire.The EN data were acquired from two Q235 carbon steel specimens placed in 0.5 mol/L H2SO4 solution and 0.1 mol/L NaCl solution.The experimental result shows that this system can achieve an accuracy of 10 pA and 10 μV,and it can be applied to on-site multi-channel EN test.展开更多
Using the quantitative error probability density method we studied the S/N ratio of alternately sampled signals digitized by a 4-channel A/D. A complete expression for the S/N ratio of a 4-channel A/D non-uniform samp...Using the quantitative error probability density method we studied the S/N ratio of alternately sampled signals digitized by a 4-channel A/D. A complete expression for the S/N ratio of a 4-channel A/D non-uniform sampling signal was deduced. First we obtained an expression for the S/N ratio of a 1-channel A/D uniform sampling signal when the sampling frequency was equal to or greater than 2 times the frequency of the sampled signal. Based on the S/N ratio of a 2-channel A/D,alternating,non-uniform sampling signal,we analyzed the distribution of quantitative error using the quantitative error probability density method and the distribution convolution formula. From this the S/N ratio expression of a 4-channel A/D sampling signal was deduced. The simulation result shows that the deduced expression is correct.展开更多
Awake monkey fMRI and diffusion MRI combined with conventional neuroscience techniques has the potential to study the structural and functional neural network. The majority of monkey fMRI and diffusion MRI experiments...Awake monkey fMRI and diffusion MRI combined with conventional neuroscience techniques has the potential to study the structural and functional neural network. The majority of monkey fMRI and diffusion MRI experiments are performed with single coils which suffer from severe EPI distortions which limit resolution. By constructing phased array coils for monkey MRI studies, gains in SNR and anatomical accuracy (i.e., reduction of EPI distortions) can be achieved using parallel imaging. The major challenges associated with constructing phased array coils for monkeys are the variation in head size and space constraints. Here, we apply phased array technology to a 4-channel phased array coil capable of improving the resolution and image quality of full brain awake monkey fMRI and diffusion MRI experiments. The phased array coil is that can adapt to different rhesus monkey head sizes (ages 4-8) and fits in the limited space provided by monkey stereotactic equipment and provides SNR gains in primary visual cortex and anatomical accuracy in conjunction with parallel imaging and improves resolution in fMRI experiments by a factor of 2 (1.25 mm to 1.0 mm isotropic) and diffusion MRI experiments by a factor of 4 (1.5 mm to 0.9 mm isotropic).展开更多
Nanocrystalline 2J4 alloy was fabricated by equal-channel angular pressing (ECAP). Microstructural evolution at different passes of ECAP and the effect of angle (φ) on the ECAP were researched. The results reveal th...Nanocrystalline 2J4 alloy was fabricated by equal-channel angular pressing (ECAP). Microstructural evolution at different passes of ECAP and the effect of angle (φ) on the ECAP were researched. The results reveal that α phase slowly turns to γ phase and follows the formation of dislocation cells in the 2J4 alloy with increasing severe plastic deformation. At last, it becomes reasonably finer bands of subgrains. The results with intersect at angle (φ) of 90° are better than that at angle (φ) of 120°. After three passes of ECAP, at angle φ of 90°, nanocrystalline microstructure can be obtained. The grain size is reduced from 30μm in the initial state to 400nm.展开更多
Nowadays,the rapid development of edge computing has driven an increasing number of deep learning applications deployed at the edge of the network,such as pedestrian and vehicle detection,to provide efficient intellig...Nowadays,the rapid development of edge computing has driven an increasing number of deep learning applications deployed at the edge of the network,such as pedestrian and vehicle detection,to provide efficient intelligent services to mobile users.However,as the accuracy requirements continue to increase,the components of deep learning models for pedestrian and vehicle detection,such as YOLOv4,become more sophisticated and the computing resources required for model training are increasing dramatically,which in turn leads to significant challenges in achieving effective deployment on resource-constrained edge devices while ensuring the high accuracy performance.For addressing this challenge,a cloud-edge collaboration-based pedestrian and vehicle detection framework is proposed in this paper,which enables sufficient training of models by utilizing the abundant computing resources in the cloud,and then deploying the well-trained models on edge devices,thus reducing the computing resource requirements for model training on edge devices.Furthermore,to reduce the size of the model deployed on edge devices,an automatic pruning method combines the convolution layer and BN layer is proposed to compress the pedestrian and vehicle detection model size.Experimental results show that the framework proposed in this paper is able to deploy the pruned model on a real edge device,Jetson TX2,with 6.72 times higher FPS.Meanwhile,the channel pruning reduces the volume and the number of parameters to 96.77%for the model,and the computing amount is reduced to 81.37%.展开更多
文摘To investigate the mechanisms of microwave induced pacemaker cell injuries, Wistar rats and the primary pacemaker cells of newborn Wistar rats were exposed to microwave at average power density of 50 mW/cm2. Slower spontaneous beating rate, intercellular Ca2＋ aggregation and cell membrane perforation were detected immediately after the exposure. Moreover, hyperpolarizationactivated cyclic nucleotide-gated cation channel 4 （HCN4） was down-regulated immediately after the exposure and up-regulated at 12 h after the exposure. In the sinoatrial node （SAN） of the rats,
基金Supported by the National Basic Research Program of China("973"Program,No.2011CB610505)National Natural Science Foundation of China(No.61240038)
文摘A portable 4-channel electrochemical noise(EN) test system with high precision was developed.The modular instrument cRIO was used as its core and the signal conditioning module included zero resistance ammeter(ZRA),fly line,screening box and shielded wire.The EN data were acquired from two Q235 carbon steel specimens placed in 0.5 mol/L H2SO4 solution and 0.1 mol/L NaCl solution.The experimental result shows that this system can achieve an accuracy of 10 pA and 10 μV,and it can be applied to on-site multi-channel EN test.
基金Projects 07KJZ11 supported by the President Fund of Xuzhou Medical School07KJB310117 by the Education Department of Jiangsu Province
文摘Using the quantitative error probability density method we studied the S/N ratio of alternately sampled signals digitized by a 4-channel A/D. A complete expression for the S/N ratio of a 4-channel A/D non-uniform sampling signal was deduced. First we obtained an expression for the S/N ratio of a 1-channel A/D uniform sampling signal when the sampling frequency was equal to or greater than 2 times the frequency of the sampled signal. Based on the S/N ratio of a 2-channel A/D,alternating,non-uniform sampling signal,we analyzed the distribution of quantitative error using the quantitative error probability density method and the distribution convolution formula. From this the S/N ratio expression of a 4-channel A/D sampling signal was deduced. The simulation result shows that the deduced expression is correct.
文摘Awake monkey fMRI and diffusion MRI combined with conventional neuroscience techniques has the potential to study the structural and functional neural network. The majority of monkey fMRI and diffusion MRI experiments are performed with single coils which suffer from severe EPI distortions which limit resolution. By constructing phased array coils for monkey MRI studies, gains in SNR and anatomical accuracy (i.e., reduction of EPI distortions) can be achieved using parallel imaging. The major challenges associated with constructing phased array coils for monkeys are the variation in head size and space constraints. Here, we apply phased array technology to a 4-channel phased array coil capable of improving the resolution and image quality of full brain awake monkey fMRI and diffusion MRI experiments. The phased array coil is that can adapt to different rhesus monkey head sizes (ages 4-8) and fits in the limited space provided by monkey stereotactic equipment and provides SNR gains in primary visual cortex and anatomical accuracy in conjunction with parallel imaging and improves resolution in fMRI experiments by a factor of 2 (1.25 mm to 1.0 mm isotropic) and diffusion MRI experiments by a factor of 4 (1.5 mm to 0.9 mm isotropic).
文摘Nanocrystalline 2J4 alloy was fabricated by equal-channel angular pressing (ECAP). Microstructural evolution at different passes of ECAP and the effect of angle (φ) on the ECAP were researched. The results reveal that α phase slowly turns to γ phase and follows the formation of dislocation cells in the 2J4 alloy with increasing severe plastic deformation. At last, it becomes reasonably finer bands of subgrains. The results with intersect at angle (φ) of 90° are better than that at angle (φ) of 120°. After three passes of ECAP, at angle φ of 90°, nanocrystalline microstructure can be obtained. The grain size is reduced from 30μm in the initial state to 400nm.
基金supported by Key-Area Research and Development Program of Guangdong Province(2021B0101420002)the Major Key Project of PCL(PCL2021A09)+3 种基金National Natural Science Foundation of China(62072187)Guangdong Major Project of Basic and Applied Basic Research(2019B030302002)Guangdong Marine Economic Development Special Fund Project(GDNRC[2022]17)Guangzhou Development Zone Science and Technology(2021GH10,2020GH10).
文摘Nowadays,the rapid development of edge computing has driven an increasing number of deep learning applications deployed at the edge of the network,such as pedestrian and vehicle detection,to provide efficient intelligent services to mobile users.However,as the accuracy requirements continue to increase,the components of deep learning models for pedestrian and vehicle detection,such as YOLOv4,become more sophisticated and the computing resources required for model training are increasing dramatically,which in turn leads to significant challenges in achieving effective deployment on resource-constrained edge devices while ensuring the high accuracy performance.For addressing this challenge,a cloud-edge collaboration-based pedestrian and vehicle detection framework is proposed in this paper,which enables sufficient training of models by utilizing the abundant computing resources in the cloud,and then deploying the well-trained models on edge devices,thus reducing the computing resource requirements for model training on edge devices.Furthermore,to reduce the size of the model deployed on edge devices,an automatic pruning method combines the convolution layer and BN layer is proposed to compress the pedestrian and vehicle detection model size.Experimental results show that the framework proposed in this paper is able to deploy the pruned model on a real edge device,Jetson TX2,with 6.72 times higher FPS.Meanwhile,the channel pruning reduces the volume and the number of parameters to 96.77%for the model,and the computing amount is reduced to 81.37%.
