Search

In this work, the Feature Pyramid Network (FPN) is proposed for improving the performance of electroencephalography (EEG) emotion recognition. Differential Entropy (DE) is extracted from each EEG channel as the basic feature. Then the feature matrix was constructed through biharmonic spline interpolation to obtain the correlation information between EEG channels. Next, the proposed FPN is applied to fuse multiscale EEG spatial information for obtaining deeper semantic features. FPN integrates the semantic information of low-level and high-level feature maps through up-sampling and lateral connections. It can reduce the semantic gap between feature maps and increase the receptive field of low-resolution feature maps. Finally, a linear Support Vector Machine (SVM) is utilized for emotion recognition. The experiment results of the proposed method have a state-of-the-art performance of valence (94.29%) and arousal (96.97%) by 70–30 cross-validation on the Database for Emotion Analysis using Physiological Signals (DEAP). And the accuracy of 97.12% was obtained on the SJTU Emotion EEG Dataset (SEED). Moreover, we carried out the LOSO CV experiment on DEAP. The accuracy of valence and arousal reached 80.38% and 82.33%, respectively.

Contains fulltext : 238698.pdf (Publisher’s version ) (Open Access), Vertebral labelling and segmentation are two fundamental tasks in an automated spine processing pipeline. Reliable and accurate processing of spine images is expected to benefit clinical decision support systems for diagnosis, surgery planning, and population-based analysis of spine and bone health. However, designing automated algorithms for spine processing is challenging predominantly due to considerable variations in anatomy and acquisition protocols and due to a severe shortage of publicly available data. Addressing these limitations, the Large Scale Vertebrae Segmentation Challenge (VerSe) was organised in conjunction with the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) in 2019 and 2020, with a call for algorithms tackling the labelling and segmentation of vertebrae. Two datasets containing a total of 374 multi-detector CT scans from 355 patients were prepared and 4505 vertebrae have individually been annotated at voxel level by a human-machine hybrid algorithm (https://osf.io/nqjyw/, https://osf.io/t98fz/). A total of 25 algorithms were benchmarked on these datasets. In this work, we present the results of this evaluation and further investigate the performance variation at the vertebra level, scan level, and different fields of view. We also evaluate the generalisability of the approaches to an implicit domain shift in data by evaluating the top-performing algorithms of one challenge iteration on data from the other iteration. The principal takeaway from VerSe: the performance of an algorithm in labelling and segmenting a spine scan hinges on its ability to correctly identify vertebrae in cases of rare anatomical variations. The VerSe content and code can be accessed at: https://github.com/anjany/verse.

A new panel-level silicon carbide (SiC) metal oxide semiconductor field effect transistor (MOSFET) power module was developed by using the fan-out and embedded chip technologies. To achieve the more effective thermal management and higher reliability under thermal cycling, a new optimization method called Ant colony optimization-back propagation neural network (ACO-BPNN) was developed for optimizing SiC modules, and contrast it with the Response Surface Method (RSM). First, the heat dissipations of SiC MOSFET with different redistribution layer (RDL) materials were simulated through the ANSYS finite element simulation. Then, the RSM was adopted to design the experiments for optimization. Third, the optimized design considering both junction temperature and thermal-mechanical stress is obtained using RSM and ACO-BPNN. The results show that: 1) compared with nano-silver, copper has a relatively good heat dissipation effect, but nano-silver has a better thermodynamic performance, and 2) ACO-BPNN can provide more accurate optimization results without having to construct a fitness function like RSM. After optimization, the thermal management and the thermal-mechanical stress can be improved by about 3% and 11%, respectively., Electronic Components, Technology and Materials

There is a large body of literature linking anatomic and geometric characteristics of kidney tumors to perioperative and oncologic outcomes. Semantic segmentation of these tumors and their host kidneys is a promising tool for quantitatively characterizing these lesions, but its adoption is limited due to the manual effort required to produce high-quality 3D segmentations of these structures. Recently, methods based on deep learning have shown excellent results in automatic 3D segmentation, but they require large datasets for training, and there remains little consensus on which methods perform best. The 2019 Kidney and Kidney Tumor Segmentation challenge (KiTS19) was a competition held in conjunction with the 2019 International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) which sought to address these issues and stimulate progress on this automatic segmentation problem. A training set of 210 cross sectional CT images with kidney tumors was publicly released with corresponding semantic segmentation masks. 106 teams from five continents used this data to develop automated systems to predict the true segmentation masks on a test set of 90 CT images for which the corresponding ground truth segmentations were kept private. These predictions were scored and ranked according to their average Sørensen-Dice coefficient between the kidney and tumor across all 90 cases. The winning team achieved a Dice of 0.974 for kidney and 0.851 for tumor, approaching the inter-annotator performance on kidney (0.983) but falling short on tumor (0.923). This challenge has now entered an "open leaderboard" phase where it serves as a challenging benchmark in 3D semantic segmentation.

International climate policy debate has been struggling to define an agreeable principle for just distribution of the burdens of reducing greenhouse gases. By the integrated assessment modeling, we investigate climate change mitigation pathways compatible with the Paris Agreement goals under different carbon budget allocation principles, namely, cost-effectiveness, equality, and grandfathering for 11 world regions. The results indicate that regional carbon quotas remain similar under equality and grandfathering principles, except for regions with a high population projection, i.e., Sub-Saharan Africa and South Asia, or regions with a high level of current emissions, such as North America. Irrespective of the burden-sharing principles, most world regions must reach carbon neutrality before 2070 to meet the 2°C climate targets. Moreover, carbon emissions reduced by carbon capture and storage applications are found to peak around 2050. In general, the share of fossil fuels in primary energy is larger under the larger carbon quota, which will also increase carbon capture and storage applications in fossil fuels. The average global investment needs for mitigating climate change are the lowest under the cost-effectiveness principle, and the marginal abatement cost is the same for all regions. For each region, as the carbon quota decreases the marginal abatement cost increases, while the energy investment changes independently from the carbon quota. These insights can inform global climate change negotiations and help policymakers formulate timelines for carbon neutrality, energy transition, and technology deployment and investment portfolios.

The development of electric vehicles can help reduce the use of fossil fuels and further mitigate greenhouse gas (GHG) emissions. A comprehensive method is proposed for estimating the GHG mitigation potential of passenger battery electric vehicles (PBEVs) continentally and globally, considering the electricity consumption, energy transition, and other main influencing factors. The future energy transition is considered to change the electricity generation structures, improve the share of clean power, and further increase the GHG mitigation potential of electric vehicles based on a life-cycle analysis. An uncertainty analysis is performed to investigate the main influencing factors, such as energy intensity, battery size, autonomous vehicles, charging infrastructure, carpooling and ridesharing, and the development of other competitive vehicles on the electricity consumption and mitigation potential. Results show that the global stock of PBEVs will reach 1.2 × 108 and 1.0 × 109 vehicles in 2030 and 2050, respectively, and the global electricity consumption will be 250–480 TWh in 2030 and 1140–2840 TWh in 2050, equivalent of saving 3.4-6.5 × 108 barrel of gasoline in 2030 and 1.6-3.9 × 109 barrel of gasoline in 2050. The calculated global GHG mitigation potentials of PBEVs are 40–215 Mt CO2e and 340–1380 Mt CO2e in 2030 and 2050, respectively, which will accelerate the decarbonization transition in the transport sector and help achieve the global temperature control goals in the Paris Agreement and sustainable development goals.

In this article, a microchannel thermal management system (MTMS) with the two-phase flow using the refrigerant R1234yf with low global warming potential is presented. The thermal test vehicles (TTVs) were made of either single or multiple thermal test chips embedded in the substrates, which were then attached to the MTMS. The system included two identical aluminum microchannel heat sinks (MHSs) connected in series in the cooling loop, which also consisted of a gas flowmeter, a miniature compressor, a condenser, a throttling device, and accessory measurement components. The experimental results showed that the thermal management system could dissipate a heat flux of 526 W/cm2 while maintaining the junction temperature below 120 °C. For SiC mosfet with a higher junction temperature, e.g., 175 °C, the current system is expected to dissipate a heat flux as high as about 750 W/cm2. The effects of the rotational speed of the compressor, the opening of the throttling device, TTV layout on MHS, and a downstream heater on the cooling performance of the system were analyzed in detail. The study shows that the present thermal management with a two-phase flow system is a promising cooling technology for the high heat flux SiC devices., Accepted Auhtor Manuscript, Electronic Components, Technology and Materials

In this article, a novel fan-out panel-level printed circuit board (PCB)-embedded package for phase-leg silicon carbide (SiC) metal-oxide-semiconductor field-effect transistor (MOSFET) power module is presented. Electro-thermo-mechanical co-design was conducted, and the maximum package parasitic inductance was found to be about 1.24 nH at 100 kHz. Compared with wire-bonded packages, the parasitic inductances of the PCB-embedded package decreased at least by 87.6%. Compared with blind via structure, the thermal resistance of the proposed blind block structure reduced at most by about 26%, and the stress of the SiC MOSFETs decreased by about 45.2%. Then, a novel PCB-embedded packaging process was developed, and three key packaging processes were analyzed. Furthermore, effect of PCB-embedded package on static characterization of SiC MOSFET was analyzed, and it was found that: 1) Output current of PCB-embedded package was decreased under a certain gate-source voltage compared to SiC die; 2) Miller capacitance of SiC MOSFET was increased thanks to parasitic capacitance induced by package; and 3) compared with SiC die, nonflat miller plateau of the PCB-embedded package extends, and as drain-source voltage increases, the nonflat miller plateau extends. Lastly, switching characteristics of the PCB-embedded package and TO-247 package were compared. The results show that the PCB-embedded package has smaller parasitic inductances., Electronic Components, Technology and Materials

SiC devices are promising for outperforming Si counterparts in high-frequency applications due to its superior material properties. Conventional wirebonded packaging scheme has been one of the most preferred package structures for power modules. However, the technique limits the performance of a SiC power module due to parasitic inductance and heat dissipation issues that are inherent with aluminum wires. In this article, low parasitic inductance and high-efficient cooling interconnection techniques for Si power modules, which are the foundation of packaging methods of SiC ones, are reviewed first. Then, attempts on developing packaging techniques for SiC power modules are thoroughly overviewed. Finally, scientific challenges in the packaging of SiC power module are summarized., Electronic Components, Technology and Materials, DC systems, Energy conversion & Storage, Electrical Sustainable Energy

Electronic Components, Technology and Materials

In order to quantify the contribution of the mitigation strategies, an extended Kaya identity has been proposed in this paper for decomposing the various factors that influence the CO2 emission. To this end, we provided a detailed decomposition of the carbon intensity and energy intensity, which enables the quantification of clean energy development and electrification. The logarithmic mean divisia index (LMDI) has been applied to the historical data to quantify the contributions of the various factors affecting the CO2 emissions. Further, the global energy interconnection (GEI) scenario has been introduced for providing a systematic solution to meet the 2°C goal of the Paris Agreement. By combining LMDI with the scenario analysis, the mitigation potential of the various factors for CO2 emission has been analyzed. Results from the historical data indicate that economic development and population growth contribute the most to the increase in CO2 emissions, whereas improvement in the power generation efficiency predominantly helps in emission reduction. A numerical analysis, performed for obtaining the projected future carbon emissions, suggests that clean energy development and electrification are the top two factors that can decrease CO2 emissions, thus showing their great potential for mitigation in the future. Moreover, the carbon capture and storage technology serves as an important supplementary mitigation method.

Background

Methods

Results

Conclusions

Level of evidence

In this paper, a novel fan-out panel-level printed circuit board (PCB) embedded package technology for silicon carbide (SiC) MOSFET power module is presented to address parasitic inductances, heat dissipation, and reliability issues that are inherent with aluminum wires used in conventional packaging scheme. To withstand high temperature beyond 175 °C and high voltage over 1.2 kV and improve thermomechanical reliability of the fan-out panel-level PCB embedded SiC power module, bismaleimide-triazine (BT) laminate and prepreg with high-temperature stability, high dielectric strength, coefficient of thermal expansion (CTE) matching with SiC, and high T-g are selected as PCB embedded package materials. Then, high-temperature stabilities, dielectric breakdown strength, and thermomechanical performances of the embedded materials are characterized. The experimental results show that the PCB embedded materials can withstand high temperature beyond 200 °C and a high voltage above 1.2 kV. T-g is as high as over 260 °C, and CTE is matching with SiC. Besides, in order to provide one guideline for the high-temperature and high-pressure laminating process during the PCB embedded SiC MOSFETs packaging, cure kinetics of BT prepreg are analyzed. The results show that 1-h curing time at 280 °C curing temperature and 2-h curing time at 210 °C curing temperature can ensure the full cure of the BT prepreg., Accepted author manuscript, DC systems, Energy conversion & Storage, Electronic Components, Technology and Materials, ESE Programmes

Cellular networks are now facing severe traffic overload problems due to the explosive growth of mobile data traffic. One of the promising solutions is to offload part of the traffic through WiFi. In this paper, we investigate an oligopoly offloading market, where several mobile virtual network operators (MVNOs) compete to serve end users using network infrastructures leased from the host mobile network operator (MNO). First, we study the competitive interaction among multiple MVNOs considering the overload problems of the offloading market. Then, we formulate the interaction as a noncooperative inventory game, where each MVNO determines the amount of cellular traffic provided to end users (named as the traffic inventory). Particularly, we investigate two different behavior patterns of the MVNOs known as Cournot and Stackelberg models. Then, we analyze and derive the existence and uniqueness of the equilibrium in each inventory game. Furthermore, algorithms are designed to achieve the equilibriums. Based on these analyses, we find the optimal inventory strategy for these competing MVNOs. Finally, simulation results demonstrate the interactions among the MNO, MVNOs, and end users in the offloading market.