Search

To analyze the risk factors affecting prognosis of children with hemophagocytic lymphohistiocytosis (HLH).The clinical manifestations and laboratory data of 143 HLH children who met the HLH-2004 diagnostic criteria in Shenzhen Children's Hospital from January 2009 to May 2017 were retrospectively analyzed, and the independent factors affecting prognosis were also analyzed.The median age of 143 HLH children was 1.9 (0.1-14.3) years old, and the median follow-up time was 6.7 years (1 day - 11.9 years). The overall survival rate of 1 month, 1 year, and 10 years was (87.4±5.5)%, (81.1±6.5)%, and (81.1±6.5)%, respectively. The deaths occurred within 1 year after onset. Multivariate analysis showed that central nervous system (CNS) involvement (P=0.047), low hemoglobin (P=0.002), prolonged activated partial thromboplastin time (APTT) (P0.001), high triglyceride (P=0.005) were all the independent risk factors affecting survival of the children. Receiver operating characteristic curve indicated that APTT (AUC=0.753, P0.001) was more valuable than other risk factors in predicting death of the children. The cut-off value of APTT was 56.6 s, and the sensitivity and specificity of which was 55.6% and 89.7%, respectively.Hypohemoglobinemia, prolonged APTT, hypertriglyceridemia, and CNS involvement the risk factors affecting prognosis of HLH, and prolonged APTT shows a strong predictive value for death.儿童噬血细胞性淋巴组织细胞增生症预后相关因素分析.分析噬血细胞性淋巴组织细胞增生症（HLH）患儿的预后相关危险因素.回顾性分析深圳市儿童医院2009年1月至2017年5月143例符合HLH-2004诊断标准的HLH患儿临床表现及实验室检查，分析影响患儿预后的独立危险因素.患儿的中位年龄为1.9（0.1-14.3）岁，中位随访时间为6.7年（1天-11.9年）；1个月、1年和10年总生存率分别为（87.4±5.5）%、（81.1±6.5）%和（81.1±6.5）%，死亡集中发生在发病后1年内；多因素分析结果显示，中枢神经系统受累（P=0.047）、血红蛋白降低（P=0.002）、活化部分凝血活酶时间延长（P0.001）、高甘油三酯（P=0.005）均是影响患儿生存的独立危险因素；受试者工作特征曲线分析提示，活化部分凝血活酶时间（AUC=0.753，P0.001）较其他危险因素预测死亡的价值更大，其预测死亡的截断值为56.6 s，灵敏度和特异度分别为55.6%和89.7.血红蛋白降低、活化部分凝血活酶时间延长、高甘油三酯及中枢神经系统受累是影响HLH预后的危险因素，其中活化部分凝血活酶时间延长对死亡具有较强的预测价值.

Point set is a flexible and lightweight representation widely used for 3D deep learning. However, their discrete nature prevents them from representing continuous and fine geometry, posing a major issue for learning-based shape generation. In this work, we turn the discrete point sets into smooth surfaces by introducing the well-known implicit moving least-squares (IMLS) surface formulation, which naturally defines locally implicit functions on point sets. We incorporate IMLS surface generation into deep neural networks for inheriting both the flexibility of point sets and the high quality of implicit surfaces. Our IMLSNet predicts an octree structure as a scaffold for generating MLS points where needed and characterizes shape geometry with learned local priors. Furthermore, our implicit function evaluation is independent of the neural network once the MLS points are predicted, thus enabling fast runtime evaluation. Our experiments on 3D object reconstruction demonstrate that IMLSNets outperform state-of-the-art learning-based methods in terms of reconstruction quality and computational efficiency. Extensive ablation tests also validate our network design and loss functions., Accepted by CVPR 2021, Code: https://github.com/Andy97/DeepMLS

The performance of seed laser source will have a great impact on the output pulse quality of the master oscillator power amplifier (MOPA) system. Here, we report the influence of the soliton vector features on the pulse uniformity in an all-fiber MOPA configuration. We demonstrated that, although the intensity of the dissipative soliton seed source is uniform, the intensity of the output pulse train from the MOPA will show fluctuations if the soliton polarization state is not locked, i.e., polarization-rotating vector soliton. It is found that the appearance of non-uniform pulse intensity is caused by the combination of the polarization instability and the polarization dependent gain in the polarization maintaining fiber amplifier. The obtained results indicate that the polarization evolution of seed pulsed source is critical to generation of stable pulse train from the MOPA, and also might be useful to the communities dealing with solitons and high-power pulsed fiber lasers.

The effects of polymorphic form and particle size of SiO2 fillers on the dielectric, mechanical and thermal properties of SiO2–Polyetheretherketone (SiO2–PEEK) composites were investigated in this paper. Strong low frequency (2 fillers. All the composites obtained in this case demonstrate the dielectric permittivities of [Formula: see text] at high frequencies. The crystalline [Formula: see text]-cristobalite filled composite exhibits lower dielectric loss and mechanical strength, but larger thermal expansion coefficient and thermal conductivity, compared with the similar particle sized amorphous SiO2 filled one. The crystalline [Formula: see text]-quartz filled composite demonstrates the lowest mechanical strength and highest dielectric loss. An increase in particle size of the spherical fused silica fillers decreases the dielectric loss, while increases the thermal conductivity of the composite. The flexural strength of the composite reaches the maximum value of 113 MPa when the particle size of spherical SiO2 filler is [Formula: see text]m. Particle packing by combining optimal amounts of differently sized spherical fused silica fillers leads to a substantial improvement of mechanical strength (153MPa) coupled with reasonable dielectric and thermal properties due to the synergic effect (dielectric permittivity ([Formula: see text] = 3.35, dielectric loss (tan[Formula: see text] @10 GHz, thermal conductivity ([Formula: see text] = 0.74 W/m*k ([Formula: see text]C), coefficient of thermal expansion ([Formula: see text]C and relative density ([Formula: see text]) = 99.72%).

3D geometric features constitute rich details of polygonal meshes. Their analysis and editing can lead to vivid appearance of shapes and better understanding of the underlying geometry for shape processing and analysis. Traditional mesh smoothing techniques mainly focus on noise filtering and they cannot distinguish different scales of features well, even mixing them up. We present an efficient method to process different scale geometric features based on a novel rolling-guidance normal filter. Given a 3D mesh, our method iteratively applies a joint bilateral filter to face normals at a specified scale, which empirically smooths small-scale geometric features while preserving large-scale features. Our method recovers the mesh from the filtered face normals by a modified Poisson-based gradient deformation that yields better surface quality than existing methods. We demonstrate the effectiveness and superiority of our method on a series of geometry processing tasks, including geometry texture removal and enhancement, coating transfer, mesh segmentation and level-of-detail meshing.

Light absorption plays a key role in photovoltaic devices, especially in thin-film CdTe-based solar cells. Here we study the enhancement of optical absorption in a ultrathin CdTe layer by embedding it into optical magnetic metamaterials. In a periodic metal-CdS-CdTe-metal sandwiched structure, the absorption of transverse magnetic polarized light has been significantly enhanced. The maximum enhancement is around 3 times, which is attributed to the strong local field enhancement in the CdTe thin film. Importantly, due to the coexistence of large electric resonance and magnetic resonance in magnetic metamaterials, the enhancement of light absorption has been obtained in a broadband from 400 nm to 800 nm, which is almost 1.5 times of that of the conventional CdTe-based photovoltaic device and covers the strongest part of the solar spectrum. Moreover, such absorption enhancement is also valid for a wide range of incident angles. We believe that our finding can lead to a variety of important applications in ultrathin CdTe-based solar cell technology.

Here we demonstrate the control of magnetic dipole spontaneous emission at yellow light by polarization-independent hexagonally arrayed nanorods magnetic metamaterials. By embedding a magnetic dipole into a polarization-independent magnetic metamaterial consisting of hexagonal arrays of paired silver nanorods, the radiative emission enhancement and the Purcell factor around 590 nm have been dramatically increased 44 times for a significant general far-field emission enhancement and 57 times for a maximum general magnetic dipole near-field emission enhancement, respectively. Moreover, the polarization-independent enhancements are found to be a robust variation of the dipole’s positions and structure geometries, showing nice fabrication tolerance for practical applications.