Approximating radiance fields with discretized volumetric grids is one of promising directions for improving NeRFs, represented by methods like DVGO, Plenoxels and TensoRF, which achieve super-fast training convergence and real-time rendering. However, these methods typically require a tremendous storage overhead, costing up to hundreds of megabytes of disk space and runtime memory for a single scene. We address this issue in this paper by introducing a simple yet effective framework, called vector quantized radiance fields (VQRF), for compressing these volume-grid-based radiance fields. We first present a robust and adaptive metric for estimating redundancy in grid models and performing voxel pruning by better exploring intermediate outputs of volumetric rendering. A trainable vector quantization is further proposed to improve the compactness of grid models. In combination with an efficient joint tuning strategy and post-processing, our method can achieve a compression ratio of 100× by reducing the overall model size to 1 MB with negligible loss on visual quality. Extensive experiments demonstrate that the proposed framework is capable of achieving unrivaled performance and well generalization across multiple methods with distinct volumetric structures, facilitating the wide use of volumetric radiance fields methods in real-world applications. Code is available at https://github.com/AlgoHunt/VQRF.

The increasing demand and limited natural resources for rare precious metal palladium render its separation and recovery from secondary sources such as acidic nuclear liquid waste greatly interesting. Herein, a novel Et-TolDAPhen extraction resin was prepared by impregnating Amberlite XAD-7 resin with a phenanthroline-derived diamide extractant N,N & PRIME;-diethyl-N,N & PRIME;-ditolyl-2,9-diamide-1,10-phenanthroline (Et-Tol-DAPhen), and the adsorption behavior of this new extraction resin towards Pd2+ in HNO3 solution was investigated through the batch and column experiments. The Et-Tol-DAPhen extraction resin exhibited strong adsorbability, high adsorption capacity, excellent selectivity and good reusability for Pd2+ in highly acidic HNO3 solution. In addition, the pseudo-second-order and Langmuir models were applied to understand the adsorption process of Pd2+ on Et-Tol-DAPhen extraction resin, respectively. It demonstrated the characteristics of a uniform singlelayer chemisorption. The maximum uptake capacity could reach as high as 74.6 mg g-1 for Pd2+ in 3.0 mol/L HNO3 solution which is much higher than most adsorbent materials of palladium reported before. Furthermore, Et-Tol-DAPhen extraction resin could selectively uptake Pd2+ from a simulated acidic nuclear waste solution containing 20 different kinds of ions with the separation factor SFPd/M values as high as 100 to 10,000. Moreover, although the adsorption efficiency had a slight decrease after 5 times of adsorption-desorption process, it still remained at a high level. This new extraction resin displayed a potential application prospect in the field of separation and recovery of Pd2+ from acidic nuclear liquid waste.

Reconstructing neural radiance fields with explicit volumetric representations, demonstrated by Plenoxels, has shown remarkable advantages on training and rendering efficiency, while grid-based representations typically induce considerable overhead for storage and transmission. In this work, we present a simple and effective framework for pursuing compact radiance fields from the perspective of compression methodology. By exploiting intrinsic properties exhibiting in grid models, a non-uniform compression stem is developed to significantly reduce model complexity and a novel parameterized module, named Neural Codebook, is introduced for better encoding high-frequency details specific to per-scene models via a fast optimization. Our approach can achieve over 40 × reduction on grid model storage with competitive rendering quality. In addition, the method can achieve real-time rendering speed with 180 fps, realizing significant advantage on storage cost compared to real-time rendering methods.

In this paper, we present a novel and effective framework, named 4K-NeRF, to pursue high fidelity view synthesis on the challenging scenarios of ultra high resolutions, building on the methodology of neural radiance fields (NeRF). The rendering procedure of NeRF-based methods typically relies on a pixel-wise manner in which rays (or pixels) are treated independently on both training and inference phases, limiting its representational ability on describing subtle details, especially when lifting to a extremely high resolution. We address the issue by exploring ray correlation to enhance high-frequency details recovery. Particularly, we use the 3D-aware encoder to model geometric information effectively in a lower resolution space and recover fine details through the 3D-aware decoder, conditioned on ray features and depths estimated by the encoder. Joint training with patch-based sampling further facilitates our method incorporating the supervision from perception oriented regularization beyond pixel-wise loss. Benefiting from the use of geometry-aware local context, our method can significantly boost rendering quality on high-frequency details compared with modern NeRF methods, and achieve the state-of-the-art visual quality on 4K ultra-high-resolution scenarios. Code Available at \url{https://github.com/frozoul/4K-NeRF}

We present an explicit-grid based method for efficiently reconstructing streaming radiance fields for novel view synthesis of real world dynamic scenes. Instead of training a single model that combines all the frames, we formulate the dynamic modeling problem with an incremental learning paradigm in which per-frame model difference is trained to complement the adaption of a base model on the current frame. By exploiting the simple yet effective tuning strategy with narrow bands, the proposed method realizes a feasible framework for handling video sequences on-the-fly with high training efficiency. The storage overhead induced by using explicit grid representations can be significantly reduced through the use of model difference based compression. We also introduce an efficient strategy to further accelerate model optimization for each frame. Experiments on challenging video sequences demonstrate that our approach is capable of achieving a training speed of 15 seconds per-frame with competitive rendering quality, which attains $1000 \times$ speedup over the state-of-the-art implicit methods. Code is available at https://github.com/AlgoHunt/StreamRF.

Effective and selective separation of technetium from acidic nuclear liquid waste is highly desirable for partitioning and transmutation but is of significant challenge. Highly efficient extraction of pertechnetate can be achieved by taking H-bonding and electrostatic interaction combined strategy. Base on this strategy, an amine-amide ligand NTAamide(n-Oct) was employed to extract TcO4- in HNO3 solution. Using n-dodecane as a diluent, NTAamide(n-Oct) demonstrated excellent extractability and good selectivity toward TcO4- with a rapid extraction equilibrium that could be reached in less than 1 min. Its maximal loading capacity for TcO4- was almost 100 times as much as that of traditional amine extractant Aliquat-336 nitrate. Meanwhile, TcO4- could be efficiently stripped from the loaded organic phase by (NH4)2CO3 solution. Slope analysis indicated the formation of a 1:1 complex of NTAamide(n-Oct) with TcO4-. The extraction conformed to the anion exchange extraction model, as confirmed by analyses of single-crystal X-ray diffraction, 1H NMR titration, FTIR, and ESI-MS.

Under the trend of electronic invoicing recording, replacing labour with automatic key information recognition will help regulators improve their work efficiency. Due to the low image quality of some electronic invoices, the existing text recognition methods still have some drawbacks. On the one hand, some factors such as lighting, tilt, scale, etc., may affect the definition of the photograph in the process of electronic invoicing, resulted in blurring the key information and increasing the difficulty of text recognition. On the other hand, the templates of electronic invoices are often fixed, and the existing methods lack the use of prior knowledge, which increases the computational cost. Given the above facts, this paper proposes a novel knowledge-based key information detection and recognition method for electronic invoices in complex scenes. In the training stage, we simulate complex scenes by using data augmentation techniques such as adding random noise, colour jitter, horizontal lines, and random rotation to improve the accuracy of text recognition. In the modelling stage, the proposed method makes full use of prior knowledge to obtain key information slices to improve the efficiency of recognition processing. Finally, in practice, the effectiveness of the method proposed in this paper is proved.

Purpose Acute kidney injury (AKI) is a serious complication of sepsis and is characterized by inflammatory response. MicroRNA-210 host gene (MIR210HG) is upregulated in human proximal tubular epithelial cells under treatment of inflammatory cytokines. This study aimed to explore the role of MIR210HG in sepsis-induced AKI. Materials and Methods Cell viability was detected by a cell counting kit 8 assay. The levels of proinflammatory cytokines were detected by enzyme-linked immunosorbent assay kits. The protein levels of p65, IκBα, and p-IκBα were examined by western blot analysis. The nuclear translocation of nuclear factor kappa B (NF-κB) was detected by immunofluorescence assay. The histological changes of kidneys were analyzed by hematoxylin and eosin staining assay. Results Lipopolysaccharide (LPS) treatment significantly inhibited cell viability and increased productions of proinflammatory cytokines in proximal tubular epithelial cells (HKC-8). Additionally, MIR210HG levels in HKC-8 cells were increased by LPS treatment. MIR210HG silencing inhibited the LPS-induced cell inflammatory response. MIR210HG activated the NF-κB signaling pathway by promoting the phosphorylation of IκBα and nuclear translocation of p65. Rescue assays revealed that the MIR210HG-induced increase of cytokines levels and decline of cell viability were rescued by QNZ treatment. Knockdown of MIR210HG decreased blood urea nitrogen, serum creatinine, and proinflammatory cytokine levels in AKI rats. Moreover, the knockdown of MIR210HG protected against AKI-induced histological changes of kidneys in rats. Conclusion MIR210HG promotes sepsis-induced inflammatory response of HKC-8 cells by activating the NF-κB signaling pathway. This novel discovery may be helpful for the improvement of sepsis-induced AKI.

Even though 'Exercise is Medicine' has been emphasized as a crucial element of healthy lifestyle, limited evidence has been provided for its effect in treating patient with acute and life-threatening disease, such as the COVID-19 pandemic. After the eruption of COVID-19 at the beginning of 2020, shelter hospitals were established in Wuhan. Special treatments were employed to stop the unprecedented disease and physical exercise was also recognized highly by doctors in the shelter hospitals for its effectiveness in promoting the rehabilitation of patients. PURPOSE: The study aims at unveiling the rationale of including physical exercise as vital part of the rehabilitation plan of patients suffering COVID-19, examining its effectiveness, i.e., what works and how it works, and influence to others through investigating the real-life experience of doctor, nurses in the shelter hospitals. It is also sought to construct a theoretical roadmap for the implementation and role of 'Exercise is Medicine' health care system in the treatment of acute and and life-threatening disease through the experience learnt from the case of shelter hospitals in Wuhan. RESEARCH METHOD: Two types of qualitative data are collected for the study. 10 interviews were conducted with two doctors and five nurses who worked in the shelter hospitals and two local residents and one front-line journalist, who were all in the city in the pandemic circumstance. In addition, reports concerning shelter hospital and the treatment in specific, and the pandemic in general, by Chinese media were also collected in order to provide background information of the pandemic. RESULT: As the result of the analysis of qualitative data, Figure 1 illustrates the roadmap and operational mechanism of the implementation of 'Exercise is Medicine' health care system in shelter hospitals in Wuhan. And physical exercise works effectively as a supplementary means of medical treatment in promoting the well-being of patients, who are in the late stage of their treatment. Along with its positive impact in promoting cardiopulmonary function of patients, it also shows significant benefits for patients in dealing with psychological issues, for instance, as an effective means of releasing mental pressure and regaining confidence.

Deep convolutional neural networks (CNN) have achieved astonishing results in a large variety of applications. However, using these models on mobile or embedded devices is difficult due to the limited memory and computation resources. Recently, the inverted residual block becomes the dominating solution for the architecture design of compact CNNs. In this work, we comprehensively investigated the existing design concepts, rethink the functional characteristics of two pointwise convolutions in the inverted residuals. We propose a novel design, called asymmetrical bottlenecks. Precisely, we adjust the first pointwise convolution dimension, enrich the information flow by feature reuse, and migrate saved computations to the second pointwise convolution. Doing so we can further improve the accuracy without increasing the computation overhead. The asymmetrical bottlenecks can be adopted as a dropin replacement for the existing CNN blocks. We can thus create AsymmNet by easily stack those blocks according to proper depth and width conditions. Extensive experiments demonstrate that our proposed block design is more beneficial than the original inverted residual bottlenecks for mobile networks, especially useful for those ultralight CNNs within the regime of <220M MAdds. Code is available at https://github.com/Spark001/AsymmNet

The highly efficient removal of tetracycline (TC) from an aqueous solution was accomplished by using the raw shrimp shell waste (SSW) as an environmentally friendly adsorbent. The SSW without any treatment removed TC more efficiently than the SSW after being treated with HCl and NaOH solutions. The SSW was characterized using nitrogen adsorption-desorption isotherms, scanning electron microscopy alongside energy-dispersive X-ray spectroscopy, Fourier transform infrared spectroscopy, a thermogravimetric-derivative thermogravimetry analyzer, and a ζ-potential analyzer. The maximum adsorption capacity of 400 mg/L SSW was 229.98 mg/g for 36 h at 55 °C. Both the Langmuir isotherm model and the pseudo-second-order kinetic model well described the experimental data. According to the values of the Gibbs free energy and enthalpy changes, the TC adsorption by SSW proved to be spontaneous and endothermic. The TC adsorption process was controlled by intraparticle diffusion and liquid film diffusion.

Objective: This study aimed to evaluate the therapeutic response of hepatocellular carcinoma (HCC) after transcatheter arterial chemoembolization (TACE) with diffusion kurtosis imaging (DKI). Methods: Forty-three patients with fifty-nine hepatic cancer nodules were recruited for this study. All patients were treated by TACE. Magnetic resonance imaging (MRI) and DKI (b=0, 800, 1,500, 2,000mm2/s) were performed before and one month after initiating TACE. Patients were classified as either progressing groups or non-progressing groups. Mean kurtosis (MK), mean diffusion (MD), and apparent diffusion coefficient (ADC) values of the tumor tissue were analyzed. Results: Twenty-three HCCs were classified as progressing groups, and thirty-six HCCs were non-progressing groups. After TACE, the values of MD and ADC in non-progressing groups (1.92±0.36×10-3mm2/s, 1.36±0.23×10-3mm2/s) were greater than progressing groups (1.44±0.32× 10-3mm2/s, 1.10±0.23×10-3mm2/s), however, the MK values in non-progressing groups (0.47±0.12) were lower than progressing groups (0.72±0.14). The MK values of tumor among non-progressing patients decreased one month after TACE (0.47±0.12) relative to the preoperative values (0.71±0.12) (P<0.05). In the non-progressing groups, the MD and ADC values of tumor after TACE (1.92±0.36×10-3mm2/s, 1.36±0.23×10-3mm2/s) became higher than their preoperative values (1.44±0.35×10-3mm2/s, 1.09±0.22×10-3mm2/s) (P<0.05). In the progressing groups, the MK, MD, and ADC values of tumor after TACE remained similar before TACE (P>0.05). The sensitivity, specificity, and AUC of the ROC curve for the assessment of HCC progress after TACE by MK (85.2%, 97.5%, and 0.95, respectively) were greater than by ADC (78.6%, 66.5%, and 0.75, respectively) and MD (76.2%, 64.3%, and 0.71, respectively). Conclusions: DKI for assessing the therapeutic response of TACE in HCC shows great promise. MK is more advantageous in the assessment of HCC progress after TACE.

目的 分析正常子宫肌层、高强度聚焦超声(HIFU)治疗子宫腺肌病(AM)前后增强型T2加权血管成像(ESWAN)参数值,探讨ESWAN在AM的诊断及评估HIFU治疗的应用价值.方法 分析44例经HIFU治疗AM患者及40例志愿者磁共振平扫,增强及ESWAN序列资料.记录AM治疗前后及子宫肌层幅度值、相位值、R2值和T2值,分析ESWAN不同参数在AM中的诊断价值及评估HIFU治疗的效果.结果 子宫肌层幅度值、相位值、R2值和T2值分别为(1652.36±425.66)、(0.032±0.060)、(22.35±5.17)、(43.26±9.36).AM患者HIFU治疗前后幅度值分别为(1242.47±346.81)和(1402.22±279.98),相位值为(0.014±0.044)和(0.020±0.039)、R2值为(26.30±2.59)和(23.15±2.40)、T2值为(39.60±4.23)和(41.71±4.72).AM患者HIFU治疗前的幅度值和T2值均低于子宫肌层(P<0.05、0.027),R2值降低(P<0.05);相位值低于子宫肌层(P=0.108),但差异无统计学意义.HIFU治疗后,幅度值和T2值升高(P<0.05),R2值降低(P≤0.05);相位值升高(P=0.249),但差异无统计学意义.结论 ESWAN在一定程度反映子宫腺肌病病变区及肌层间的信号差异,同时能够反应AM患者HIFU治疗后的组织学变化.

Colon cancer is the most commonly diagnosed malignancy and the leading cause of cancer deaths worldwide. As well as lifestyle, genetic and epigenetic changes are key factors that influence the risk of colon cancer. However, the impact of epigenetic alterations in non-coding RNAs and their consequences in colon cancer have not been fully characterized. We detected differential methylation sites (DMSs) in long non-coding RNA (lncRNA) promoters and identified lncRNA expression quantitative trait methylations (lncQTMs) by association tests. To investigate how transcription factor (TF) binding was affected by DNA methylation, we characterized the occurrence of known TFs among DMSs collected from the MEME suite. We further combined methylome and transcriptome data to construct TF-methylation-lncRNA relationships. To study the role of lncRNAs in drug response, we used pharmacological and lncRNA profiles from the Cancer Cell Line Encyclopedia (CCLE) and investigated the association between lncRNAs and drug activity. We also used combinations of TF-methylation-lncRNA relationships to stratify patient survival using a risk model. DNA methylation sites displayed global hyper-methylation in lncRNA promoters and tended to have negative relationships with the corresponding lncRNAs. Negative lncQTMs located near transcription start sites (TSSs) had more significant correlations with the corresponding lncRNAs. Some lncRNAs found to be mediated by the interplay between DNA methylation and TFs were previously identified as markers for colon cancer. We also found that the ELF1-cg05372727- LINC00460 relationship were prognostic signatures for colon cancer. These findings suggest that lncRNAs mediated by the interplay between DNA methylation and TFs are promising predictors of drug response, and that combined TF-methylation-lncRNA can serve as a prognostic signature for colon cancer.

In this study, we isolated 20 novel microsatellites loci associated with the growth of Hexagrammos otakii (Fat greenling) by using 2b-RAD sequencing method. The characteristics of 20 microsatellite loci were amplified in 105 H. otakii individuals which came from four different groups and tested by Capillary electrophoresis. The number of allele of the 20 microsatellite loci ranged from 8 to 26 with an average of 19.95. PIC value ranged from 0.1659 to 0.9227 with an average of 0.7555. Observed and expected heterozygosity varied from 0.0667 to 0.8571 and from 0.1710 to 0.9314, respectively. Highly polymorphic characteristics were observed in each microsatellite loci of the study except HO6. Further research showed that 10 microsatellite loci had transferability in Hexagrammos agrammus. In this study, high polymorphism and genetic diversity using the 15 polymorphic microsatellite loci suggest that they are suitable for investigating the fine-scale population structure, genetic relationship and further evaluating the artificial reproduction strategy of H. otakii.

To clarify evolutionary characteristics, phylogenetic relationships as well as species identification of C. okamurae, we determined the cpDNA sequence of Caulerpa okamurae using de novo sequencing in the present study. The cpDNA of C. okamurae was 148,274 bp in length, and it lacked the inverted repeat commonly found in vascular green plants. The cpDNA of C. okamurae was highly compact with a gene density of 71.7%. Moreover, it was an AT-rich genome (65.5%) consisting 76 protein-coding genes (PCGs), 27 transfer RNA (tRNA) genes, three ribosomal RNA (rRNA) genes, 32 putative open reading frames (ORFs) and six introns. Additionally, the six introns were annotated in six genes as follows: psbA, rpoB, ftsH, psbD, atpF and cysA. The overall base composition of its cpDNA was 65.46% for AT. A total of 56 genes were encoded on the light strand, while all the other 50 chloroplast genes were encoded on the heavy strand. All of the PCGs had ATG as their start codon and employed TAA, TGA or TAG as their termination codon. Phylogenetic analyses suggested that the complete cpDNA sequence of C. okamurae fell in the Chlorophyta, Ulvophyceae, Bryopsidales, and Caulerpaceae and more resembled the cpDNAs of C. racemosa, C. cliftonii voucher and Tydemania expeditionis. Taken together, our data offered useful information for the studies of C.okamurae on evolutionary characteristics, phylogenetic relationships as well as species identification.

PURPOSE: The study aims at unveiling key influential factors of the EIM in China and investigating the relationship among the factors through employing Interpretative Structural Model (ISM). METHODS: Twelve semi-structured interviews were conducted with experts (9male,3 female) with medical, sport and government background. Matrix questionnaires were also completed by the interviewees. Data Analysis:1) Interpretative Structural Model (ISM) was designed following a consultation process with expert group, and key factors were selected to construct adjacency matrix 1)A=[aij]=1,SiRSj(R indicates Si and Sj related); A=[aij]=0, SiR’Sj(R indicates Si and Sj related); 2) Calculate the reachability matrix by Boolean Calculation,M=(A+I)λ+1=(A+I)λ≠(A+I)λ-1≠··(A+I)2≠(A+I);3)Decompose reachability matrix, L1={Si|R(Si)∩A(Si)=R(Si)=R(Si),i=0,1,2,k},and then delete rows and columns corresponding to elements in L1, the reachability matrix is obtained; 4) Establish multi-level directed graph and analysis SIM. RESULTS: The first level contains R(S5)∩A(S5)=R(S5);R(S11)∩A(S11)=R(S11),so the factors are{S5,S11⋯},and then delete the row and column S5,S11⋯, in the matrix, we can get the second level of the factors, the second level is {S7,S8,S13},And so on, The third level is {S4,S6,S12},The fourth level is {S1,S3,S9};The fifth level is {S2,S10}. These factors of Exercise is Medicine in China can be divided into three layers by ISM analysis, including the surface factors, the middle factors and the decisive factors. CONCLUSIONS: As decisive factors, the policy system and economic environment, which are the significantly influential to the policy related to EIM, could be regarded as important contextual factors for the design and implementation of EIM policy. The middle factors mainly focused on Governance capability and collaboration between sports and medical administration. Relatively, surface factors contains more but pays less attention.