A Universal Self-Propagating Synthesis of Aluminum-Based Oxyhalide Solid-State Electrolytes

Inorganic solid-state electrolytes (SSEs) play a vital role in high-energy all-solid-state batteries (ASSBs). However, the current method of SSE preparation usually involves high-energy mechanical ball milling and/or a high-temperature annealing process, which is not suitable for practical application. Here, a facile strategy is developed to realize the scalable synthesis of cost-effective aluminum-based oxyhalide SSEs, which involves a self-propagating method by the exothermic reaction of the raw materials. This strategy enables the synthesis of various aluminum-based oxyhalide SSEs with tunable components and high ionic conductivities (over 10-3 S cm-1 at 25 °C) for different cations (Li+, Na+, Ag+). It is elucidated that the amorphous matrix, which mainly consists of various oxidized chloroaluminate species that provide numerous sites for smooth ion migration, is actually the key factor for the achieved high conductivities. The application of these aluminum-based oxyhalide SSEs synthesized by our approach further pushes forward their practical application considering their easy synthesis, low cost, and low weight that ensures high-energy-density ASSBs.

Effect of resistance circuit training on comprehensive health indicators in older adults: a systematic review and meta-analysis

The aging process leads to the degeneration of body structure and function. The objective of this study is to conduct a systematic review and meta-analysis of the effects of resistance circuit training (RCT) on comprehensive health indicators of older adults. PubMed, Embase, and Web of Science were searched until August 2023. Primary outcomes were body composition, muscle strength, cardiorespiratory endurance, blood pressure, and functional autonomy. Muscle function and exercise intensity subgroups were analyzed. RCT reduces body fat (MD = - 5.39 kg, 95% CI - 10.48 to - 0.29), BMI (MD = - 1.22, 95% CI - 2.17 to - 0.26), and body weight (MD = - 1.28 kg, 95% CI - 1.78 to - 0.78), and increases lean body mass (MD = 1.42 kg, 95% CI 0.83-2.01) in older adults. It improves upper limb strength (SMD = 2.09, 95% CI 1.7-2.48), lower limb strength (SMD = 2.03, 95% CI 1.56-2.51), cardiorespiratory endurance (MD = 94 m, 95% CI 25.69-162.67), and functional autonomy (MD = - 1.35, 95% CI - 1.73 to - 0.96). High-intensity RCT benefits BMI and body weight, while low-intensity exercise reduces blood pressure. RCT improves muscle function in push, pull, hip, and knee movements in older adults. RCT improves body composition, muscle strength, cardiorespiratory endurance, blood pressure, and functional autonomy in older adults. High-intensity training is superior for body composition, while moderate to low intensity training is more effective for lowering blood pressure.

Ethylene Vinyl Alcohol Copolymer Nanofibrous Cation Exchange Chromatographic Membranes with a Gradient Porous Structure for Lysozyme Separation

Lysozyme, a common antimicrobial agent, is widely used in the food, biopharmaceutical, chemical, and medicine fields. Rapid and effective isolation of lysozymes is an everlasting topic. In this work, ethylene vinyl alcohol (EVOH) copolymer nanofibrous membranes with a gradient porous structure used for lysozyme adsorption were prepared through layer-by-layer nanofiber wet-laying and a cost-efficient ultraviolet (UV)-assisted graft-modification method, where benzophenone was used as an initiator and 2-acrylamide-2-methylpropanesulfonic acid as a modifying monomer. As indicated in the Fourier Transform Infrared (FTIR) and X-ray photoelectric energy spectrometer (XPS) investigation, sulfonic acid groups were introduced on the surface of the modified nanofibrous membrane, which possessed the ability to adsorb lysozyme. Compared with membranes with homogenous porous structures, membranes with a gradient porous structure present higher static (335 mg/g) and dynamic adsorption capacities (216.3 mg/g). Meanwhile, the adsorption capacity remained high after five cycles of the adsorption-desorption process. The results can be attributed to the gradient porous structure rather than the highest porosity and specific surface area. This suggests that the membrane with comprehensive separation performance can be designed from the view of the transmembrane porous structure, which is of significance for the development of next-generation advanced chromatographic membranes.

Hypoxia treatment and resistance training alters microRNA profiling in rats skeletal muscle

MicroRNAs (miRNAs) may play a crucial regulatory role in the process of muscle atrophy induced by high-altitude hypoxia and its amelioration through resistance training. However, research in this aspect is still lacking. Therefore, this study aimed to employ miRNA microarray analysis to investigate the expression profile of miRNAs in skeletal muscle from an animal model of hypoxia-induced muscle atrophy and resistance training aimed at mitigating muscle atrophy. The study utilized a simulated hypoxic environment (oxygen concentration at 11.2%) to induce muscle atrophy and established a rat model of resistance training using ladder climbing, with a total intervention period of 4 weeks. The miRNA expression profile revealed 9 differentially expressed miRNAs influenced by hypoxia (e.g., miR-341, miR-32-5p, miR-465-5p) and 14 differentially expressed miRNAs influenced by resistance training under hypoxic conditions (e.g., miR-338-5p, miR-203a-3p, miR-92b-3p) (∣log2(FC)∣ ≥ 1.5, p < 0.05). The differentially expressed miRNAs were found to target genes involved in muscle protein synthesis and degradation (such as Utrn, mdm2, eIF4E), biological processes (such as negative regulation of transcription from RNA polymerase II promoter, regulation of transcription, DNA-dependent), and signaling pathways (such as Wnt signaling pathway, MAPK signaling pathway, ubiquitin-mediated proteolysis, mTOR signaling pathway). This study provides a foundation for understanding and further exploring the molecular mechanisms underlying hypoxia-induced rats muscle atrophy and the mitigation of atrophy through resistance training.

Predictive value of cyst/tumor volume ratio of pituitary adenoma for tumor cell proliferation

BACKGROUND

MRI has been widely used to predict the preoperative proliferative potential of pituitary adenoma (PA). However, the relationship between the cyst/tumor volume ratio (C/T ratio) and the proliferative potential of PA has not been reported. Herein, we determined the predictive value of the C/T ratio of PA for tumor cell proliferation.

METHODS

The clinical data of 72 patients with PA and cystic change on MRI were retrospectively analyzed. PA volume, cyst volume, and C/T ratio were calculated. The corresponding intraoperative specimens were collected. Immunohistochemistry and hematoxylin-eosin staining were performed to evaluate the Ki67 index and nuclear atypia. Patients were categorized according to the Ki67 index (< 3% and ≥ 3%) and nuclear atypia (absence and presence). Univariate and multivariate analyses were used to identify the significant predictors of the Ki67 index and nuclear atypia. The receiver operating characteristic curve assessed the prediction ability of the significant predictors.

RESULTS

Larger tumor volumes, smaller cyst volumes, and lower C/T ratios were found in patients with higher Ki67 indexes and those with nuclear atypia (P < 0.05). C/T ratio was an independent predictor of the Ki67 index (odds ratio = 0.010, 95% confidence interval = 0.000-0.462) and nuclear atypia (odds ratio = 0.010, 95% confidence interval = 0.000-0.250). The predictive value of the C/T ratio did not differ significantly from that of tumor volume (P > 0.05) but was better than that of cyst volume (P < 0.05). The area under the curve of the C/T ratio for predicting the Ki67 index and nuclear atypia was larger than that for predicting cyst volume and tumor volume.

CONCLUSIONS

C/T ratios can be used to predict PA tumor proliferation preoperatively. Our findings may facilitate the selection of surgery timing and the efficacy evaluation of surgery.

HIRI-ViT: Scaling Vision Transformer with High Resolution Inputs

The hybrid deep models of Vision Transformer (ViT) and Convolution Neural Network (CNN) have emerged as a powerful class of backbones for vision tasks. Scaling up the input resolution of such hybrid backbones naturally strengthes model capacity, but inevitably suffers from heavy computational cost that scales quadratically. Instead, we present a new hybrid backbone with HIgh-Resolution Inputs (namely HIRI-ViT), that upgrades prevalent four-stage ViT to five-stage ViT tailored for high-resolution inputs. HIRI-ViT is built upon the seminal idea of decomposing the typical CNN operations into two parallel CNN branches in a cost-efficient manner. One high-resolution branch directly takes primary high-resolution features as inputs, but uses less convolution operations. The other low-resolution branch first performs down-sampling and then utilizes more convolution operations over such low-resolution features. Experiments on both recognition task (ImageNet-1K dataset) and dense prediction tasks (COCO and ADE20K datasets) demonstrate the superiority of HIRI-ViT. More remarkably, under comparable computational cost (  ∼ 5.0 GFLOPs), HIRI-ViT achieves to-date the best published Top-1 accuracy of 84.3% on ImageNet with 448×448 inputs, which absolutely improves 83.4% of iFormer-S by 0.9% with 224×224 inputs.

Comparison of the Efficacy and Safety of Perioperative Immunochemotherapeutic Strategies for Resectable Non-small Cell Lung Cancer: a Systematic Review and Network Meta-analysis

AIMS

The aim of this network meta-analysis was to elucidate the efficacy and safety of various immune checkpoint inhibitors (ICIs) used in combination with chemotherapy for the treatment of non-small cell lung cancer (NSCLC).

MATERIALS AND METHODS

Data from randomised controlled trials comparing perioperative ICI-chemotherapy and chemotherapy alone were acquired from the EMBASE, Web of Science, Cochrane Library databases, PubMed, and meeting abstracts from inception until August 2023. The endpoints for this analysis were pathological complete response, event-free survival and treatment-related adverse events of any grade or adverse events of grade 3 or higher.

RESULTS

In total, six randomised controlled trials with 2538 NSCLC patients were selected for this network meta-analysis. Compared with other ICIs, toripalimab + chemotherapy demonstrated increased pathological complete response rates and prolonged event-free survival in NSCLC. In patients with negative/low PD-L1 expression or squamous cell pathology, toripalimab + chemotherapy was the most effective regimen. In contrast, nivolumab + chemotherapy was preferable for patients with high PD-L1 expression or non-squamous cell pathology. Among the analysed regimens, toripalimab + chemotherapy presented the highest risk of adverse events of any grade, whereas nivolumab + chemotherapy showed the highest risk of grade 3-4 adverse events. Conversely, durvalumab + chemotherapy exhibited the lowest risk of grade 3-4 adverse events.

CONCLUSIONS

Among the evaluated perioperative immunochemotherapy regimens, toripalimab + chemotherapy indicated a significantly increased survival benefit for most resectable NSCLC patients. However, for high PD-L1 expression and non-squamous NSCLC patients, nivolumab + chemotherapy provided the most potent outcomes. Perioperative durvalumab + chemotherapy is a relatively safe treatment. The findings of this investigation are expected to assist clinicians in making informed decisions among promising treatment options.

A Genotype-Phenotype Model for Predicting Resistance Training Effects on Leg Press Performance

This study develops a comprehensive genotype-phenotype model for predicting the effects of resistance training on leg press performance. A cohort of physically inactive adults (N=193) underwent 12 weeks of resistance training, and measurements of maximum isokinetic leg press peak force, muscle mass, and thickness were taken before and after the intervention. Whole-genome genotyping was performed, and genome-wide association analysis identified 85 novel SNPs significantly associated with changes in leg press strength after training. A prediction model was constructed using stepwise linear regression, incorporating seven lead SNPs that explained 40.4% of the training effect variance. The polygenic score showed a significant positive correlation with changes in leg press strength. By integrating genomic markers and phenotypic indicators, the comprehensive prediction model explained 75.4% of the variance in the training effect. Additionally, five SNPs were found to potentially impact muscle contraction, metabolism, growth, and development through their association with REACTOME pathways. Individual responses to resistance training varied, with changes in leg press strength ranging from -55.83% to 151.20%. The study highlights the importance of genetic factors in predicting training outcomes and provides insights into the potential biological functions underlying resistance training effects. The comprehensive model offers valuable guidance for personalized fitness programs based on individual genetic profiles and phenotypic characteristics.

Structural regulation of halide superionic conductors for all-solid-state lithium batteries

Metal halide solid-state electrolytes have gained widespread attention due to their high ionic conductivities, wide electrochemical stability windows, and good compatibility with oxide cathode materials. The exploration of highly ionic conductive halide electrolytes is actively ongoing. Thus, understanding the relationship between composition and crystal structure can be a critical guide for designing better halide electrolytes, which still remains obscure for reliable prediction. Here we show that the cationic polarization factor, which describes the geometric and ionic conditions, is effective in predicting the stacking structure of halide electrolytes formation. By supplementing this principle with rational design and preparation of more than 10 lithium halide electrolytes with high conductivity over 10-3 S cm-1 at 25 °C, we establish that there should be a variety of promising halide electrolytes that have yet to be discovered and developed. This methodology may enable the systematic screening of various potential halide electrolytes and demonstrate an approach to the design of halide electrolytes with superionic conductivity beyond the structure and stability predictions.

Case Report: Hemorrhagic Fever with Renal Syndrome Complicated by Bilateral Subdural Hematoma

Hemorrhagic fever with renal syndrome (HFRS) is an acute, natural focal disease worldwide. Bilateral subdural hematoma (BSH) is a rare occurrence in patients with HFRS. A 51-year-old man was admitted with fever, headache, lower back pain, and reduced urine volume. The patient was diagnosed with HFRS accompanied by BSH, as evidenced by IgM and IgG antibodies for hantavirus that were positive, and abnormal blood test results and computed tomographic head scan. He recovered and was discharged after symptomatic treatment. Hemorrhagic fever with renal syndrome might present rare clinical manifestations with BSH. The early identification of this condition is crucial to an improved prognosis.

The Universal Super Cation-Conductivity in Multiple-cation Mixed Chloride Solid-State Electrolytes

As exciting candidates for next-generation energy storage, all-solid-state lithium batteries (ASSLBs) are highly dependent on advanced solid-state electrolytes (SSEs). Here, using cost-effective LaCl3 and CeCl3 lattice (UCl3 -type structure) as the host and further combined with a multiple-cation mixed strategy, we report a series of UCl3 -type SSEs with high room-temperature ionic conductivities over 10-3  S cm-1 and good compatibility with high-voltage oxide cathodes. The intrinsic large-size hexagonal one-dimensional channels and highly disordered amorphous phase induced by multi-metal cation species are believed to trigger fast multiple ionic conductions of Li+ , Na+ , K+ , Cu+ , and Ag+ . The UCl3 -type SSEs enable a stable prototype ASSLB capable of over 3000 cycles and high reversibility at -30 °C. Further exploration of the brand-new multiple-cation mixed chlorides is likely to lead to the development of advanced halide SSEs suitable for ASSLBs with high energy density.

N-Halamine-Based Polypropylene Melt-Blown Nonwoven Fabric with Superhydrophilicity and Antibacterial Properties for Face Masks

Polypropylene melt-blown nonwoven fabric (PP MNF) masks can effectively block pathogens in the environment from entering the human body. However, the adhesion of surviving pathogens to masks poses a risk of human infection. Thus, embedding safe and efficient antibacterial materials is the key to solving pathogen infection. In this study, stable chlorinated poly(methacrylamide-N,N'-methylenebisacrylamide) polypropylene melt-blown nonwoven fabrics (PP-P(MAA-MBAA)-Cl MNFs) have been fabricated by a simple UV cross-link and chlorination process, and the active chlorine content can reach 3500 ppm. The PP-P(MAA-MBAA)-Cl MNFs show excellent hydrophilic and antibacterial properties. The PP-P(MAA-MBAA)-Cl MNFs could kill all bacteria (both Escherichia coli and Staphylococcus aureus) with only 5 min of contact. Therefore, incorporating PP-P(MAA-MBAA)-Cl MNF as a hydrophilic antimicrobial layer into a four-layer PP-based mask holds great potential for enhancing protection and comfort.

Selective leaching process for efficient and rapid recycling of spent lithium iron phosphate batteries

With the continuous development of new energy vehicles, the number of decommissioned lithium iron phosphate (LiFePO4) batteries has been constantly increasing. Therefore, it is necessary to recover metal from spent LiFePO4 batteries due to the high potential for environmental protection and high resource value. In this study, sodium persulfate (Na2S2O8) was selected as the oxidant to regulate and control the oxidation state and proton activity of the leaching solution through its high oxidizing ability. Selective recovery of lithium from LiFePO4 batteries was achieved by oxidizing LiFePO4 to iron phosphate (FePO4) during the leaching process. This paper reports an extensive investigation of the effects of various factors, including the acid concentration, initial volume fraction of the oxidant, reaction temperature, solid-liquid ratio, and reaction time, on lithium leaching. Li+ reached a high leaching rate of 93.3% within 5 minutes even at a low concentration of sulphuric acid (H2SO4), and high-purity lithium carbonate (Li2CO3) was obtained through impurity removal and precipitation reactions. In addition, the leaching mechanism was analysed by both X-ray diffraction and X-ray photoelectron spectroscopy characterization. The results show that the obtained high lithium-ion (Li+) leaching efficiency and fast Li+ leaching time can be ascribed to the superior oxidizing properties of Na2S2O8 and the stability of the crystal structure of LiFePO4 during the oxidative leaching process. The adopted method has significant advantages in terms of safety, efficiency and environmental protection, which are conducive to the sustainable development of lithium batteries.

Genotype-Phenotype Models Predicting V̇O 2max Response to High-Intensity Interval Training in Physically Inactive Chinese

PURPOSE

This study aimed to analyze the interindividual differences of the maximal oxygen uptake (V̇O 2max ) response to 12 wk of high-intensity interval training (HIIT), and the genotype-phenotype models were constructed to predict the effect of HIIT on V̇O 2max .

METHODS

A total of 228 physically inactive adults who completed a 12-wk HIIT were analyzed. A genome-wide association study (GWAS) was conducted to identify genetic variants associated with the V̇O 2max response. Nonresponders, responders, and the highest training responders were defined as the effect sizes (ES) <0.2, ≥0.2, and ≥0.8, respectively. We generated polygenic predictor score (PPS) using lead variants and constructed a predictive model for V̇O 2max response based on a linear stepwise regression analysis.

RESULTS

The V̇O 2max increased significantly after HIIT (~14%, P < 0.001), but with interindividual differences (-7.8 to 17.9 mL·kg -1 ·min -1 ). In 27% of participants, the V̇O 2max showed no improvement. We identified one genetic locus near the γ-aminobutyric acid type A receptor subunit beta 3 gene ( GABRB3 , rs17116985) associated with V̇O 2max response at the genome-wide significance level ( P < 5 × 10 -8 ), and an additional nine single nucleotide polymorphisms (SNPs) at the suggestive significance level ( P < 1 × 10 -5 ). The SNPs rs474377, rs9365605, and rs17116985, respectively, explained 11%, 9%, and 6.2% of variance in V̇O 2max response. The 13 SNPs ( P < 1 × 10 -5 ) were found on chromosome 6 (position: 148209316-148223568). Individuals with a PPS greater than 1.757 had the highest response, and those with a PPS lower than -3.712 were nonresponders. The PPS, baseline V̇O 2max , sex, and body mass explained 56.4% of the variance in the V̇O 2max response; the major predictor was the PPS, which explained 39.4% of the variance.

CONCLUSIONS

The PPS, baseline V̇O 2max , sex, and body mass could explain the variance in V̇O 2max response. Individuals who had a PPS greater than 1.757 had the highest training response after 12 wk of HIIT. Genetic variants in a region on chromosome 6, especially the sterile alpha motif domain containing 5 gene ( SAMD5 ), which had been explored influencing angiogenesis, might have a potential role in the V̇O 2max response.

Unusual Hydrogenation Reactivities of a Thiolate-Bridged Dicobalt μ-Nitride Featuring a Bent {CoIII-N-CoIII} Core

Transition metal nitrides have received considerable attention owing to their crucial roles in nitrogen fixation and nitrogen atom transfer reactions. Compared to the early and middle transition metals, it is much more challenging to access late transition metal nitrides, especially cobalt in group 9. So far, only a handful of cobalt nitrides have been reported; consequently, their hydrogenation reactivity is largely unexplored. Herein, we present a structurally and spectroscopically well-characterized thiolate-bridged dicobalt μ-nitride [Cp*CoIII(μ-SAd)(μ-N)CoIIICp*] (2) featuring a bent {CoIII(μ-N)CoIII} core. Remarkably, complex 2 can realize not only direct hydrogenation of nitride to amide but also stepwise N-H bond formation from nitride to ammonia. Specifically, 2 can facilely activate dihydrogen (H2) at mild conditions to generate a dicobalt μ-amide [Cp*CoII(μ-SAd)(μ-NH2)CoIICp*] (4) via an unusual mechanism of two-electron oxidation of H2 as proposed by computational studies; in the presence of protons (H+) and electrons, nitride 2 can convert to dicobalt μ-imide [Cp*CoIII(μ-SAd)(μ-NH)CoIIICp*][BPh4] (3[BPh4]) and to CoIICoII μ-amide 4, and finally release ammonia. In contrast to 2, the only other structurally characterized dicobalt μ-nitride Na(THF)4{[(ketguan)CoIII(N3)]2(μ-N)} (ketguan = [(tBu2CN)C(NDipp)2]-, Dipp = 2,6-diisopropylphenyl) (e) that possesses a linear {CoIII(μ-N)CoIII} moiety cannot directly react with H2 or H+. Further in-depth electronic structure analyses shed light on how the varying geometries of the {CoIII(μ-N)CoIII} moieties in 2 and e, bent vs linear, impart their disparate reactivities.

MRI features of pituitary adenoma apoplexy and their relationship with hypoxia, proliferation, and pathology

OBJECTIVE

We aim to study the MRI features of pituitary adenoma (PA) apoplexy and their relationship with hypoxia, proliferation, and pathology.

METHODS

Sixty-seven patients with MRI signs of PA apoplexy were selected. According to the MRI signs, they were divided into the parenchymal group and the cystic group. The parenchymal group had a low signal area on T2WI without cyst >2 mm and this area was not significantly enhanced on the corresponding TW1 enhancement. The cystic group had a cyst >2 mm on T2WI, and the cyst showed liquid stratification on T2WI or high signal on T1WI. The relative T1WI (rT1WI) enhancement value and relative T2WI (rT2WI) value of non-apoplexy areas were measured. Protein levels of hypoxia-inducible factor-1 (HIF-1α), pyruvate dehydrogenase kinase 1 (PDK1), and Ki67 were detected with immunohistochemistry and Western blot. Nuclear morphology was observed with HE staining.

RESULTS

The rT1WI enhancement average value, rT2WI average value, Ki67 protein expression level, and the number of abnormal nuclear morphology of non-apoplexy lesions in the parenchymal group were significantly lower than those in the cystic group. The protein expression levels of HIF-1α and PDK1 in the parenchymal group were significantly higher than those in the cystic group. HIF-1α protein was positively correlated with PDK1 but negatively correlated with Ki67.

CONCLUSION

When there is PA apoplexy, the ischemia and hypoxia of the cystic group are lesser than those of the parenchymal group, but the proliferation is stronger.

Visualizing and Understanding Patch Interactions in Vision Transformer

Vision transformer (ViT) has become a leading tool in various computer vision tasks, owing to its unique self-attention mechanism that learns visual representations explicitly through cross-patch information interactions. Despite having good success, the literature seldom explores the explainability of ViT, and there is no clear picture of how the attention mechanism with respect to the correlation across comprehensive patches will impact the performance and what is the further potential. In this work, we propose a novel explainable visualization approach to analyze and interpret the crucial attention interactions among patches for ViT. Specifically, we first introduce a quantification indicator to measure the impact of patch interaction and verify such quantification on attention window design and indiscriminative patches removal. Then, we exploit the effective responsive field of each patch in ViT and devise a window-free transformer (WinfT) architecture accordingly. Extensive experiments on ImageNet demonstrate that the exquisitely designed quantitative method is shown able to facilitate ViT model learning, leading the top-1 accuracy by 4.28% at most. More remarkably, the results on downstream fine-grained recognition tasks further validate the generalization of our proposal.

CoSeg: Cognitively Inspired Unsupervised Generic Event Segmentation

Some cognitive research has discovered that humans accomplish event segmentation as a side effect of event anticipation. Inspired by this discovery, we propose a simple yet effective end-to-end self-supervised learning framework for event segmentation/boundary detection. Unlike the mainstream clustering-based methods, our framework exploits a transformer-based feature reconstruction scheme to detect event boundaries by reconstruction errors. This is consistent with the fact that humans spot new events by leveraging the deviation between their prediction and what is perceived. Thanks to their heterogeneity in semantics, the frames at boundaries are difficult to be reconstructed (generally with large reconstruction errors), which is favorable for event boundary detection. In addition, since the reconstruction occurs on the semantic feature level instead of the pixel level, we develop a temporal contrastive feature embedding (TCFE) module to learn the semantic visual representation for frame feature reconstruction (FFR). This procedure is like humans building up experiences with "long-term memory." The goal of our work is to segment generic events rather than localize some specific ones. We focus on achieving accurate event boundaries. As a result, we adopt the F1 score (Precision/Recall) as our primary evaluation metric for a fair comparison with previous approaches. Meanwhile, we also calculate the conventional frame-based mean over frames (MoF) and intersection over union (IoU) metric. We thoroughly benchmark our work on four publicly available datasets and demonstrate much better results. The source code is available at https://github.com/wang3702/CoSeg.

Sheet-on-sheet ZnIn2S4@RGO-modified separators with abundant sulfur vacancies for high-performance Li-S batteries

A novel sheet-on-sheet architecture with abundant sulfur vacancies (Vs) is designed by in situ growth of flake-like ZnIn2S4 on the reduced graphene oxide (Vs-ZIS@RGO) surface, which serves as a functional layer on the separators for high-performance lithium-sulfur batteries (LSBs). Benefiting from the sheet-on-sheet architecture, the separators exhibit rapid ionic/electronic transfer, which is capable of supporting fast redox reactions. The vertically ordered ZnIn2S4 shortens the diffusion pathways of lithium-ions and the irregularly curved nanosheets expose more active sites to effectively anchor lithium polysulfides (LiPSs). More importantly, the introduction of Vs adjusts the surface or interface electronic structure of ZnIn2S4, enhancing the chemical affinity to LiPSs while accelerating conversion reaction kinetics of LiPSs. As expected, the batteries with Vs-ZIS@RGO modified separators exhibit an initial discharge capacity of 1067 mA h g-1 at 0.5C. Even at 1C, the excellent long cycle stability (710 mA h g-1 over 500 cycles) with an ultra-low decay rate of 0.055% per cycle is also attained. This work proposes a strategy of designing the sheet-on-sheet structure with rich sulfur vacancies, which provides a new perspective to rationally devise durable and efficient LSBs.

Exercise Therapy for Knee Osteoarthritis: A Systematic Review and Network Meta-analysis

Background

Exercise is an effective nonpharmaceutical therapy for knee osteoarthritis (KOA).

Purpose

To identify the most effective type of exercise therapy for KOA with regard to pain, stiffness, joint function, and quality of life.

Study Design

Systematic review; Level of evidence, 3.

Methods

The PubMed, Web of Science, Embase, and Cochrane Library databases were searched, from inception to April 4, 2022. Included were randomized controlled trials that assessed the efficacy on KOA among 5 different exercise therapy groups (aquatic exercise [AE], stationary cycling [CY], resistance training [RT], traditional exercise [TC], and yoga [YG]) and compared with the control group. Outcomes among the groups were assessed with the Western Ontario and McMaster University Osteoarthritis Index (WOMAC), 6-minute walk test (6-MWT), visual analog scale (VAS) for pain, and Knee injury and Osteoarthritis Outcome Score (KOOS); weighted mean differences (WMDs) and 95% confidence intervals were calculated. Network meta-analyses comparing outcomes between all groups and with controls were performed, and group rankings were calculated using the surface under the cumulative ranking curve (SUCRA).

Results

A total of 39 studies (N = 2646 participants) were included. Most of the studies failed to blind participants and researchers, resulting in a high risk of performance bias. Significantly worse WOMAC-Pain scores were seen in controls compared with all exercise interventions except AE (WMD [95% CI]: CY, -4.45 [-5.69 to -3.20]; RT, -4.28 [-5.48 to -3.07]; TC, -4.20 [-5.37 to -3.04]; and YG, -0.57 [-1.04 to -1.04]), and worse scores were seen in controls compared with YG regarding WOMAC-Stiffness (WMD, -1.40 [95% CI, -2.45 to -0.34]) and WOMAC-Function (WMD, -0.49 [95% CI, -0.95 to -0.02]). According to the SUCRA, CY was the most effective for improving WOMAC-Pain (80.8%) and 6-MWT (76.1%); YG was most effective for improving WOMAC-Stiffness (90.6%), WOMAC-Function (77.4%), KOOS-Activities of Daily Living (72.0%), and KOOS-Quality of Life (79.1%); AE was the most effective regarding VAS pain (77.2%) and KOOS-Pain (64.0%); and RT was the most effective regarding KOOS-Symptoms (84.5%).

Conclusion

All 5 types of exercise were able to ameliorate KOA. AE (for pain relief) and YG (for joint stiffness, limited knee function, and quality of life) were the most effective approaches, followed by RT, CY, and TC.

Dual Vision Transformer

Recent advances have presented several strategies to mitigate the computations of self-attention mechanism with high-resolution inputs. Many of these works consider decomposing the global self-attention procedure over image patches into regional and local feature extraction procedures that each incurs a smaller computational complexity. Despite good efficiency, these approaches seldom explore the holistic interactions among all patches, and are thus difficult to fully capture the global semantics. In this paper, we propose a novel Transformer architecture that elegantly exploits the global semantics for self-attention learning, namely Dual Vision Transformer (Dual-ViT). The new architecture incorporates a critical semantic pathway that can more efficiently compress token vectors into global semantics with reduced order of complexity. Such compressed global semantics then serve as useful prior information in learning finer local pixel level details, through another constructed pixel pathway. The semantic pathway and pixel pathway are integrated together and are jointly trained, spreading the enhanced self-attention information in parallel through both of the pathways. Dual-ViT is henceforth able to capitalize on global semantics to boost self-attention learning without compromising much computational complexity. We empirically demonstrate that Dual-ViT provides superior accuracy than SOTA Transformer architectures with comparable training complexity. Source codes are available at https://github.com/YehLi/ImageNetModel.

A Low Rank Promoting Prior for Unsupervised Contrastive Learning

Unsupervised learning is just at a tipping point where it could really take off. Among these approaches, contrastive learning has led to state-of-the-art performance. In this paper, we construct a novel probabilistic graphical model that effectively incorporates the low rank promoting prior into the framework of contrastive learning, referred to as LORAC. In contrast to the existing conventional self-supervised approaches that only considers independent learning, our hypothesis explicitly requires that all the samples belonging to the same instance class lie on the same subspace with small dimension. This heuristic poses particular joint learning constraints to reduce the degree of freedom of the problem during the search of the optimal network parameterization. Most importantly, we argue that the low rank prior employed here is not unique, and many different priors can be invoked in a similar probabilistic way, corresponding to different hypotheses about underlying truth behind the contrastive features. Empirical evidences show that the proposed algorithm clearly surpasses the state-of-the-art approaches on multiple benchmarks, including image classification, object detection, instance segmentation and keypoint detection. Code is available: https://github.com/ssl-codelab/lorac.

Augmentation Pathways Network for Visual Recognition

Data augmentation is practically helpful for visual recognition, especially at the time of data scarcity. However, such success is only limited to quite a few light augmentations (e.g., random crop, flip). Heavy augmentations are either unstable or show adverse effects during training, owing to the big gap between the original and augmented images. This paper introduces a novel network design, noted as Augmentation Pathways (AP), to systematically stabilize training on a much wider range of augmentation policies. Notably, AP tames various heavy data augmentations and stably boosts performance without a careful selection among augmentation policies. Unlike traditional single pathway, augmented images are processed in different neural paths. The main pathway handles the light augmentations, while other pathways focus on the heavier augmentations. By interacting with multiple paths in a dependent manner, the backbone network robustly learns from shared visual patterns among augmentations, and suppresses the side effect of heavy augmentations at the same time. Furthermore, we extend AP to high-order versions for high-order scenarios, demonstrating its robustness and flexibility in practical usage. Experimental results on ImageNet demonstrate the compatibility and effectiveness on a much wider range of augmentations, while consuming fewer parameters and lower computational costs at inference time.

Contextual Transformer Networks for Visual Recognition

Transformer with self-attention has led to the revolutionizing of natural language processing field, and recently inspires the emergence of Transformer-style architecture design with competitive results in numerous computer vision tasks. Nevertheless, most of existing designs directly employ self-attention over a 2D feature map to obtain the attention matrix based on pairs of isolated queries and keys at each spatial location, but leave the rich contexts among neighbor keys under-exploited. In this work, we design a novel Transformer-style module, i.e., Contextual Transformer (CoT) block, for visual recognition. Such design fully capitalizes on the contextual information among input keys to guide the learning of dynamic attention matrix and thus strengthens the capacity of visual representation. Technically, CoT block first contextually encodes input keys via a 3×3 convolution, leading to a static contextual representation of inputs. We further concatenate the encoded keys with input queries to learn the dynamic multi-head attention matrix through two consecutive 1×1 convolutions. The learnt attention matrix is multiplied by input values to achieve the dynamic contextual representation of inputs. The fusion of the static and dynamic contextual representations are finally taken as outputs. Our CoT block is appealing in the view that it can readily replace each 3×3 convolution in ResNet architectures, yielding a Transformer-style backbone named as Contextual Transformer Networks (CoTNet). Through extensive experiments over a wide range of applications (e.g., image recognition, object detection, instance segmentation, and semantic segmentation), we validate the superiority of CoTNet as a stronger backbone. Source code is available at https://github.com/JDAI-CV/CoTNet.

Self-assembly Co-doped MnO2 nanorods networks with abundant oxygen vacancies modified separators for high-performance Li-S batteries

Lithium-sulfur batteries (LSBs) are broadly considered to the most promising next-generation energy storage because of the ultrahigh theoretical energy density and cost effectiveness. However, the “shuttle effect” and sluggish conversion...

Kinetic Acceleration of Lithium Polysulfide Conversion via a Copper-Iridium Alloying Catalytic Strategy in Li-S Batteries

To solve the shuttle effect of soluble lithium polysulfides (LiPSs), a porous N-doped carbon-supported copper-iridium alloy catalyst composite (CuIr/NC) has been synthesized and served as a modified cathode sulfur host for lithium-sulfur batteries (LSBs). The metal-organic framework-derived calcined carbon frameworks build efficient conductive channels for fast ion/electron transport. Furthermore, alloying noble metals Ir with thiophilic metal Cu provides abundant active sites to effectively capture LiPSs and accelerate the catalytic conversion process, originating from modulating the surface electronic structure of the metal Cu by introducing Ir atoms to affect the 3d-orbital distribution. All of the above are strongly supported by a range of characterization studies and density functional theory calculations. Benefiting from the above advantages, the LSBs generally show satisfactory cycling performance. Apart from exhibiting a terrific initial specific capacity of 1288 mA h g^-1 at 0.2 C, they can also keep long-term cycling stability under a high current density up to 5 C together with a slow specific capacity decay ratio (0.033%) per cycle after 1000 cycles. In addition, it is worth mentioning that a high areal capacity (4.7 mA h cm^-2) with a low E/S ratio (6.2 μL mg^-1) could still be accomplished at higher sulfur loading (4.3 mg cm^-2).

A novel expansive soil hardener: performance and mechanism of immersion stability

Aiming at the existing problems of poor treatment effect and immersion stability of expansive soils, a slag soil hardener (SSH, developed by Wuhan University, China) was combined with different additives to dispose in this study. The free expansion rate, compressive strength, and immersion stability of samples were compared, and the influences of different additives, curing age, and dry density on the process and mechanism of improvement were discussed. The experimental results indicated that SSH combined with quicklime had the best improvement effect on expansive soils, in which the mass ratio of raw materials was: expansive soil/SSH/quicklime = 92/4/4, and the free expansion rate decreased from 45.90% to 4.4%, compressive strength increased from 2.53 MPa to 6.69 MPa, and there was no splitting after immersion under this ratio. FTIR spectroscopy, XPS and SEM were performed to analyze the characteristic functional groups, structural forms, and morphology of samples to study the mechanism of improvement, which showed that SSH greatly reduced the proportion of montmorillonite in the whole system and further enhanced the mechanism of ion exchange, soil particle connection, and coating protection. The research can provide theoretical reference for engineering the application of expansive soil area in rainy climate and has dual economic and environmental benefits.

A Porous Hexagonal Prism Shaped C-In2-xCoxO3 Electrocatalyst to Expedite Bidirectional Polysulfide Redox in Li-S Batteries

The shuttling behavior of soluble lithium polysulfides (LPSs) extremely restricts the practical application of lithium sulfur batteries (Li-S batteries). Herein, the hollow porous hexagonal prism shaped C-In_2-xCo_xO₃ composite is synthesized to restrain the shuttle effect and accelerate reaction kinetics of LPSs. The novel hexagonal prism porous carbon skeleton not only provides a stable physical framework for sulfur active materials but also facilitates efficient electron transferring and lithium ion diffusion. Meanwhile, the polar In_2-xCo_xO₃ is equipped with strong adsorption capacity for LPSs, which is confirmed by density functional theory (DFT) calculations, helping to anchor LPSs. More importantly, the doping of Co regulates the electronic structure environment of In₂O₃, expedites the electron transmission, and bidirectionally improves the catalytic conversion ability of LPSs and nucleation-decomposition of Li₂S. Benefiting from the above advantages, the electrochemical performance of Li-S batteries has been greatly enhanced. Therefore, the C-In_2-xCo_xO₃ cathode presents a good rate performance, which exhibits a low-capacity fading rate of 0.052% per cycle over 800 cycles at 5 C. Especially, even under a high sulfur loading of 4.8 mg cm^-2, the initial specific capacity is as high as 903 mAh g^-1, together with a superior capacity retention of 85.6% after 600 cycles at 0.5 C.

3D Cascade RCNN: High Quality Object Detection in Point Clouds

Recent progress on 2D object detection has featured Cascade RCNN, which capitalizes on a sequence of cascade detectors to progressively improve proposal quality, towards high-quality object detection. However, there has not been evidence in support of building such cascade structures for 3D object detection, a challenging detection scenario with highly sparse LiDAR point clouds. In this work, we present a simple yet effective cascade architecture, named 3D Cascade RCNN, that allocates multiple detectors based on the voxelized point clouds in a cascade paradigm, pursuing higher quality 3D object detector progressively. Furthermore, we quantitatively define the sparsity level of the points within 3D bounding box of each object as the point completeness score, which is exploited as the task weight for each proposal to guide the learning of each stage detector. The spirit behind is to assign higher weights for high-quality proposals with relatively complete point distribution, while down-weight the proposals with extremely sparse points that often incur noise during training. This design of completeness-aware re-weighting elegantly upgrades the cascade paradigm to be better applicable for the sparse input data, without increasing any FLOP budgets. Through extensive experiments on both the KITTI dataset and Waymo Open Dataset, we validate the superiority of our proposed 3D Cascade RCNN, when comparing to state-of-the-art 3D object detection techniques. The source code is publicly available at https://github.com/caiqi/Cascasde-3D.

A Review of Performance Prediction Based on Machine Learning in Materials Science

With increasing demand in many areas, materials are constantly evolving. However, they still have numerous practical constraints. The rational design and discovery of new materials can create a huge technological and social impact. However, such rational design and discovery require a holistic, multi-stage design process, including the design of the material composition, material structure, material properties as well as process design and engineering. Such a complex exploration using traditional scientific methods is not only blind but also a huge waste of time and resources. Machine learning (ML), which is used across data to find correlations in material properties and understand the chemical properties of materials, is being considered a new way to explore the materials field. This paper reviews some of the major recent advances and applications of ML in the field of properties prediction of materials and discusses the key challenges and opportunities in this cross-cutting area.

Evaluation of data imputation strategies in complex, deeply-phenotyped data sets: the case of the EU-AIMS Longitudinal European Autism Project

An increasing number of large-scale multi-modal research initiatives has been conducted in the typically developing population, e.g. Dev. Cogn. Neur. 32:43-54, 2018; PLoS Med. 12(3):e1001779, 2015; Elam and Van Essen, Enc. Comp. Neur., 2013, as well as in psychiatric cohorts, e.g. Trans. Psych. 10(1):100, 2020; Mol. Psych. 19:659–667, 2014; Mol. Aut. 8:24, 2017; Eur. Child and Adol. Psych. 24(3):265–281, 2015. Missing data is a common problem in such datasets due to the difficulty of assessing multiple measures on a large number of participants. The consequences of missing data accumulate when researchers aim to integrate relationships across multiple measures. Here we aim to evaluate different imputation strategies to fill in missing values in clinical data from a large (total N = 764) and deeply phenotyped (i.e. range of clinical and cognitive instruments administered) sample of N = 453 autistic individuals and N = 311 control individuals recruited as part of the EU-AIMS Longitudinal European Autism Project (LEAP) consortium. In particular, we consider a total of 160 clinical measures divided in 15 overlapping subsets of participants. We use two simple but common univariate strategies—mean and median imputation—as well as a Round Robin regression approach involving four independent multivariate regression models including Bayesian Ridge regression, as well as several non-linear models: Decision Trees (Extra Trees., and Nearest Neighbours regression. We evaluate the models using the traditional mean square error towards removed available data, and also consider the Kullback–Leibler divergence between the observed and the imputed distributions. We show that all of the multivariate approaches tested provide a substantial improvement compared to typical univariate approaches. Further, our analyses reveal that across all 15 data-subsets tested, an Extra Trees regression approach provided the best global results. This not only allows the selection of a unique model to impute missing data for the LEAP project and delivers a fixed set of imputed clinical data to be used by researchers working with the LEAP dataset in the future, but provides more general guidelines for data imputation in large scale epidemiological studies.

Scalable 3D Honeycombed Co3O4 Modified Separators as Polysulfides Barriers for High-Performance Li-S Batteries

Lithium sulfur batteries (LSBs) are regarded as one of the most promising energy storage devices due to the high theoretical capacity and energy density. However, the shuttling lithium polysulfides (LiPSs) from the cathode and the growing lithium dendrites on the anode limit the practical application of LSBs. To overcome these challenges, a novel three-dimensional (3D) honeycombed architecture consisting of a local interconnected Co₃O₄ successfully assembled into a scalable modified layer through mutual support, which is coated on commercial separators for high-performance LSBs. On the basis of the 3D honeycombed architecture, the modified separators not only suppress effectively the "shuttle effects" but also allow for fast lithium-ions transportation. Moreover, the theoretical calculations results exhibit that the collaboration of the exposed (111) and (220) crystal planes of Co₃O₄ is able to effectively anchor LiPSs. As expected, LSBs with 3D honeycombed Co₃O₄ modified separators present a reversible specific capacity with 1007 mAh g^-1 over 100 cycles at 0.1 C. More importantly, a high reversible capacity of 808 mAh g^-1 over 300 cycles even at 1 C is also acquired with the modified separators. Therefore, this proposed strategy of 3D honeycombed architecture Co₃O₄ modified separators will give a new route to rationally devise durable and efficient LSBs.

Selective synthesis and magnetic properties of iron silicide (Fe3Si and FeSi) at low temperature

In this study, iron silicides (Fe3Si and FeSi) particles have been selectively synthesized through a solid-state reaction route. The reactions are carried out at 650 ℃ in a stainless steel...

Reversible Binding of Dinitrogen on a Thiolate-Bridged Cobalt-Ruthenium Complex Supported by a Flexible Bidentate Phosphine Ligand

A well-defined thiolate-bridged cobalt-ruthenium complex is demonstrated to reversibly bind N2 by modulation of the auxiliary phosphine ligand, which is evidenced by time-dependent 1H NMR spectroscopy at different temperatures. Notably,...

A thiolate-bridged FeIVFeIV μ-nitrido complex and its hydrogenation reactivity toward ammonia formation

Iron nitrides are key intermediates in biological nitrogen fixation and the industrial Haber-Bosch process, used to form ammonia from dinitrogen. However, the proposed successive conversion of nitride to ammonia remains elusive. In this regard, the search for well-described multi-iron nitrido model complexes and investigations on controlling their reactivity towards ammonia formation have long been of great challenge and importance. Here we report a well-defined thiolate-bridged Fe^IVFe^IV μ-nitrido complex featuring an uncommon bent Fe-N-Fe moiety. Remarkably, this complex shows excellent reactivity toward hydrogenation with H₂ at ambient conditions, forming ammonia in high yield. Combined experimental and computational studies demonstrate that a thiolate-bridged Fe^IIIFe^III μ-amido complex is a key intermediate, which is generated through an unusual two-electron oxidation of H₂. Moreover, ammonia production was also realized by treating this diiron μ-nitride with electrons and water as a proton source.

Virtual Guidance-Based Coordinated Tracking Control of Multi-Autonomous Underwater Vehicles Using Composite Neural Learning

This article proposes a virtual leader-based coordinated controller for the nonlinear multiple autonomous underwater vehicles (multi-AUVs) with the system uncertainties. To achieve the coordinated formation, a virtual AUV is set as the leader, while the desired command is designed using the relative position between each AUV and the virtual leader. The controller is designed based on the back-stepping scheme, and the online data-based learning scheme is used for uncertainty approximation. The highlight is that compared with previous learning methods which mostly focus on stability, the learning performance index is constructed using the collected online data in this article. The index is further used in the composite update law of the neural weights. The closed-loop system stability is analyzed via the Lyapunov approach. The simulation test on the five AUVs under fixed formation shows that the proposed method can achieve higher tracking performance with improved approximation accuracy.

Unpaired Person Image Generation With Semantic Parsing Transformation

In this paper, we tackle the problem of pose-guided person image generation with unpaired data, which is a challenging problem due to non-rigid spatial deformation. Instead of learning a fixed mapping directly between human bodies as previous methods, we propose a new pathway to decompose a single fixed mapping into two subtasks, namely, semantic parsing transformation and appearance generation. First, to simplify the learning for non-rigid deformation, a semantic generative network is developed to transform semantic parsing maps between different poses. Second, guided by semantic parsing maps, we render the foreground and background image, respectively. A foreground generative network learns to synthesize semantic-aware textures, and another background generative network learns to predict missing background regions caused by pose changes. Third, we enable pseudo-label training with unpaired data, and demonstrate that end-to-end training of the overall network further refines the semantic map prediction and final results accordingly. Moreover, our method is generalizable to other person image generation tasks defined on semantic maps, e.g., clothing texture transfer, controlled image manipulation, and virtual try-on. Experimental results on DeepFashion and Market-1501 datasets demonstrate the superiority of our method, especially in keeping better body shapes and clothing attributes, as well as rendering structure-coherent backgrounds.

AGRNet: Adaptive Graph Representation Learning and Reasoning for Face Parsing

Face parsing infers a pixel-wise label to each facial component, which has drawn much attention recently. Previous methods have shown their success in face parsing, which however overlook the correlation among facial components. As a matter of fact, the component-wise relationship is a critical clue in discriminating ambiguous pixels in facial area. To address this issue, we propose adaptive graph representation learning and reasoning over facial components, aiming to learn representative vertices that describe each component, exploit the component-wise relationship and thereby produce accurate parsing results against ambiguity. In particular, we devise an adaptive and differentiable graph abstraction method to represent the components on a graph via pixel-to-vertex projection under the initial condition of a predicted parsing map, where pixel features within a certain facial region are aggregated onto a vertex. Further, we explicitly incorporate the image edge as a prior in the model, which helps to discriminate edge and non-edge pixels during the projection, thus leading to refined parsing results along the edges. Then, our model learns and reasons over the relations among components by propagating information across vertices on the graph. Finally, the refined vertex features are projected back to pixel grids for the prediction of the final parsing map. To train our model, we propose a discriminative loss to penalize small distances between vertices in the feature space, which leads to distinct vertices with strong semantics. Experimental results show the superior performance of the proposed model on multiple face parsing datasets, along with the validation on the human parsing task to demonstrate the generalizability of our model.