Application of 5T glutamate chemical exchange saturation transfer imaging in brain tumors: preliminary results.

PURPOSE: Glutamate chemical exchange saturation transfer (GluCEST) is a non-invasive CEST imaging technique for detecting glutamate levels in tissues. We aimed to investigate the reproducibility of the 5T GluCEST technique in healthy volunteers and preliminarily explore its potential clinical application in patients with brain tumors. METHODS: Ten volunteers (4 males, mean age 29 years) underwent three 5T GluCEST imaging scans. The reproducibility of the three imaging GluCEST measurements was assessed using one-way repeated measures analysis of variance (ANOVA), generalized estimating equations, and linear mixed models. Twenty-eight patients with brain tumors (10 males, mean age 54 years) underwent a single GluCEST scan preoperatively, and t-tests were used to compare the differences in GluCEST values between different brain tumors. In addition, the diagnostic accuracy of GluCEST values in differentiating brain tumors was assessed using the receiver work characteristics (ROC) curve. RESULTS: The coefficients of variation of GluCEST values in healthy volunteers were less than 5% for intra-day, inter-day, and within-subjects and less than 10% for between-subjects. High-grade gliomas (HGG) had higher GluCEST values compared to low-grade gliomas (LGG) (P < 0.001). In addition, cerebellopontine angle (CPA) meningiomas had higher GluCEST values than acoustic neuromas (P < 0.001). The area under the curve (AUC) of the GluCEST value for differentiating CPA meningioma from acoustic neuroma was 0.93. CONCLUSION: 5T GluCEST images are highly reproducible in healthy brains. In addition, the 5T GluCEST technique has potential clinical applications in differentiating LGG from HGG and CPA meningiomas from acoustic neuromas.

Model test study of bridge pile horizontal bearing behavior under landslide deformation.

Pile is a common foundation on the slope, which poses a serious threat to the construction and operation if the slope deformation and causes landslide. In this study, a model device of pile foundation on landslide was independently developed by relative displacement loading between pile and soil to explore the influence of landslide deformation on pile and analysis the soil failure rule and the deformation characteristics of pile in different stages of landslide deformation, a few model tests were completed including the relative displacement between soil and pile from 1 to 17 cm, and the pile diameter and the modulus of slide bed were also considered. The results indicated that: the evolution process of landslide deformation with pile foundation on could be divided into four stages including soil compaction, cracks growth, yield stage, and failure stage; ratios of the maximum soil pressure and bending moment growth from the soil compaction stage to the cracks growth stage to the total growth in these four stages are both exceeding 60%; the soil pressure increases with the increase of pile diameter and sliding bed modulus. Therefore, it is best to effectively monitor and control the landslide in the initial soil compression stage that in soil compaction stage and methods such as increasing pile foundations or reinforcing the sliding bed can be used for protection.

Evidence of synergistic mechanisms of hepatoprotective botanical herbal preparation of Pueraria montana var. lobata and Schisandra sphenanthera.

Background: Pueraria montana var. lobata (Willd.) Maesen & S.M.Almeida ex Sanjappa & Predeep (syn. Pueraria lobata (Willd.) Ohwi) and Schisandra sphenanthera Rehder & E.H. Wilson are traditional edible and medicinal hepatoprotective botanical drugs. Studies have shown that the combination of two botanical drugs enhanced the effects of treating acute liver injury (ALI), but the synergistic effect and its action mechanisms remain unclear. This study aimed to investigate the synergistic effect and its mechanism of the combination of Pueraria montana var. lobata (Willd.) Maesen & S.M.Almeida ex Sanjappa & Predeep (syn. Pueraria lobata (Willd.) Ohwi) (PM) and Schisandra sphenanthera Rehder & E.H. Wilson (SS) in the treatment of ALI. Methods: High performance liquid chromatography (HPLC) were utilized to conduct the chemical interaction analysis. Then the synergistic effects of botanical hybrid preparation of PM-SS (BHP PM-SS) against ALI were comprehensively evaluated by the CCl4 induced ALI mice model. Afterwards, symptom-oriented network pharmacology, transcriptomics and metabolomics were applied to reveal the underlying mechanism of action. Finally, the key target genes were experimentally by RT-qPCR. Results: Chemical analysis and pharmacodynamic experiments revealed that BHP PM-SS was superior to the single botanical drug, especially at 2:3 ratio, with a better dissolution rate of active ingredients and synergistic anti-ALI effect. Integrated symptom-oriented network pharmacology combined with transcriptomics and metabolomics analyses showed that the active ingredients of BHP PM-SS could regulate Glutathione metabolism, Pyrimidine metabolism, Arginine biosynthesis and Amino acid sugar and nucleotide sugar metabolism, by acting on the targets of AKT1, TNF, EGFR, JUN, HSP90AA1 and STAT3, which could be responsible for the PI3K-AKT signaling pathway, MAPK signaling pathway and Pathway in cancer to against ALI. Conclusion: Our study has provided compelling evidence for the synergistic effect and its mechanism of the combination of BHP PM-SS, and has contributed to the development and utilization of BHP PM-SS dietary supplements.

Efficacy and Safety of Sacubitril/Valsartan in Hemodialysis Patients with Chronic Heart Failure: A Retrospective Study at a Single Center.

BACKGROUND Heart failure and end-stage renal disease often coexist, and management of heart failure can be challenging in patients during hemodialysis. Sacubitril-valsartan (SV) is the first drug to receive regulatory approval for use in patients with chronic heart failure with reduced ejection fraction (HFrEF) and New York Heart Association (NYHA) classification II, III, or IV. This study aimed to evaluate the efficacy and safety of SV for use in chronic heart failure patients on maintenance hemodialysis (MHD). MATERIAL AND METHODS From September 2021 to October 2022, 28 patients on MHD with chronic heart failure at the hemodialysis center of Shaanxi Second Provincial People's Hospital were regularly followed. During the 12-week follow-up period, all patients were administered SV at doses of 100-400 mg per day. Biochemical indicators, echocardiographic parameters, life quality scores, and adverse events were evaluated. RESULTS We enrolled 28 patients. Compared with the baseline levels, NYHA class III in these patients treated with SV was significantly decreased from 60.71% to 32.14% (P<0.05), left ventricular ejection fraction (LVEF) was significantly improved from 44.29±8.92% to 53.32±7.88% (P<0.001), the Physical Component Summary (PCS) score was significantly improved from 40.0±6.41 to 56.20±9.86 (P<0.001), and the Mental Component Summary (MCS) score was significantly improved from 39.99±6.14 to 52.59±11.0 (P<0.001). CONCLUSIONS We demonstrated that SV improved NYHA classification and LVEF values of patients on MHD with chronic heart failure and also improved their quality of life.

Preoperative prediction of MGMT promoter methylation in glioblastoma based on multiregional and multi-sequence MRI radiomics analysis.

O6-methylguanine-DNA methyltransferase (MGMT) has been demonstrated to be an important prognostic and predictive marker in glioblastoma (GBM). To establish a reliable radiomics model based on MRI data to predict the MGMT promoter methylation status of GBM. A total of 183 patients with glioblastoma were included in this retrospective study. The visually accessible Rembrandt images (VASARI) features were extracted for each patient, and a total of 14676 multi-region features were extracted from enhanced, necrotic, "non-enhanced, and edematous" areas on their multiparametric MRI. Twelve individual radiomics models were constructed based on the radiomics features from different subregions and different sequences. Four single-sequence models, three single-region models and the combined radiomics model combining all individual models were constructed. Finally, the predictive performance of adding clinical factors and VASARI characteristics was evaluated. The ComRad model combining all individual radiomics models exhibited the best performance in test set 1 and test set 2, with the area under the receiver operating characteristic curve (AUC) of 0.839 (0.709-0.963) and 0.739 (0.581-0.897), respectively. The results indicated that the radiomics model combining multi-region and multi-parametric MRI features has exhibited promising performance in predicting MGMT methylation status in GBM. The Modeling scheme that combining all individual radiomics models showed best performance among all constructed moels.

NAT10 Overexpression Promotes Tumorigenesis and Epithelial-Mesenchymal Transition Through AKT Pathway in Gastric Cancer.

BACKGROUND: N-acetyltransferase 10 (NAT10), the only RNA cytosine acetyltransferase known in humans, contributes to cancer tumorigenesis and progression. This study aims to investigate the effect of NAT10 on the malignant biological properties of gastric cancer (GC) and its underlying mechanism. METHODS: The expression and prognostic significance of NAT10 in GC were analyzed using The Cancer Genome Atlas (TCGA) and Sun Yat-sen University (SYSU) cohorts. The influence of NAT10 on the malignant biological behaviors of GC was detected by Cell Counting Kit-8 (CCK-8) assay, plate colony formation assay, 5-ethynyl-2'-deoxyuridine (EdU), Transwell migration and invasion assays, scratch wound assay, flow cytometric analysis, and animal studies. The overall level of N4 acetylcytidine (ac4C) in GC was detected by liquid chromatography with tandem mass spectrometry (LC-MS/MS). The downstream signal pathways of NAT10 were analyzed by Gene Set Enrichment Analysis (GSEA) and verified by Western blot (WB) and immunofluorescence (IF). RESULTS: The significant upregulation of NAT10 expression in GC was associated with a poor prognosis. The knockdown of NAT10 markedly suppressed GC cell proliferation, migration, invasion, and cell cycle progression. Downregulating NAT10 reduced ac4C levels and inhibited AKT phosphorylation and epithelial-mesenchymal transition (EMT) in GC. CONCLUSIONS: NAT10 functions as an oncogene and may provide a new therapeutic target in GC.

Nano co-delivery of doxorubicin and plumbagin achieves synergistic chemotherapy of hepatocellular carcinoma.

Doxorubicin (DOX) is a chemotherapy drug used for hepatocellular carcinoma (HCC) treatment, but its effectiveness can be dramatically dampened by cancer cell chemoresistance. Signal transducer and activator of transcription 3 (STAT3) is implicated with drug resistance in a range of cancers (e.g., HCC), and the STAT3 inhibition can reverse the resistance of cancer cells to chemotherapeutic drugs. In the present study, a combination regimen to improve the efficiency of DOX was provided via the STAT3 blockade using plumbagin (PLB). A poly(lactic-co-glycolic acid) decorated by polyethylene glycol and aminoethyl anisamide was produced in the present study with the hope of generating the nanoparticles for co-delivery of DOX and PLB. The resulting co-formulation suppressed the STAT3 activity and achieved the synergistic chemotherapy, which led to tumor inhibition in the mice with subcutaneous DOX-resistant HCC, without causing any toxicity. The present study reveals the synergism of DOX and PLB, and demonstrates a promising combinatorial approach for treating HCC.

Explainable machine learning for predicting 30-day readmission in acute heart failure patients.

We aimed to develop a machine-learning based predictive model to identify 30-day readmission risk in Acute heart failure (AHF) patients. In this study 2232 patients hospitalized with AHF were included. The variance inflation factor value and 5-fold cross-validation were used to select vital clinical variables. Five machine learning algorithms with good performance were applied to develop models, and the discrimination ability was comprehensively evaluated by sensitivity, specificity, and area under the ROC curve (AUC). Prediction results were illustrated by SHapley Additive exPlanations (SHAP) values. Finally, the XGBoost model performs optimally: the greatest AUC of 0.763 (0.703-0.824), highest sensitivity of 0.660, and high accuracy of 0.709. This study developed an optimal XGBoost model to predict the risk of 30-day unplanned readmission for AHF patients, which showed more significant performance compared with traditional logistic regression (LR) model.

High-risk ARGs (HRA) Chip: A high-throughput qPCR-based array for assessment of high-risk ARGs from the environment.

The global surge in antibiotic resistance genes (ARGs) presents a serious public health challenge. While methods like metagenomic analysis and qPCR arrays have been instrumental in investigating ARG distributions and dynamics, the vast diversity of ARGs often complicates effective monitoring and risk assessment. Here, we developed a High-Risk ARGs (HRA) chip based on high-capacity quantitative PCR array targeting previously identified high-risk ARGs. These ARGs are known to be prevalent in human-related environments, exhibit gene mobility, and are present in ESKAPE pathogens. The HRA chip include 101 primer sets and the 16S rRNA gene as a reference. These primer sets consist of 34 obtained from previous studies, and 67 newly designed primer sets which were validated in silico and experimentally. Absolute quantification of targeted ARGs is accomplished by generating standard curves for all ARGs with serially ten-fold diluted mixed plasmids containing targeted ARG sequences. The amplification efficiencies of all ARGs exceed 99% via plasmid template dilution tests, suggesting high reliability in quantification. The performance of HRA chip is further evaluated by practical applications in environmental samples from wastewater treatment plants (WWTPs) and soils with various land use types and fertilization regimes. The results indicate the dynamics of high-risk ARGs during wastewater treatment process, and reveal the profiles of soil high-risk ARGs which is distinct from those derived via metagenomic approaches. These findings highlight the potentials of HRA Chip in the evaluation of anthropogenic impacts on the environmental resistome with a more focused spectrum of high-risk ARGs. Overall, the HRA Chip emerges as a powerful and efficient high-throughput tool for rapid detection and quantification of high-risk ARGs, facilitating comprehensive profiling of high-risk resistomes in environmental samples which is essential for human health risk assessment of ARGs.

Research on the Welding Process and Weld Formation in Multiple Solid-Flux Cored Wires Arc Hybrid Welding Process for Q960E Ultrahigh-Strength Steel.

This paper proposes a novel welding process for ultrahigh-strength steel. The effects of welding parameters on the welding process and weld formation were studied to obtain the optimal parameter window. It was found that the metal transfer modes of solid wires were primarily determined by electrical parameters, while flux-cored wires consistently exhibited multiple droplets per pulse. The one droplet per pulse possessed better welding stability and weld formation, whereas the short-circuiting transfer or one droplet multiple pulses easily caused abnormal arc ignition that decreased welding stability, which could easily lead to a "sawtooth-shaped" weld formation or weld offset towards one side with more spatters. Thus, the electrical parameters corresponding to one droplet per pulse were identified as the optimal parameter window. Furthermore, the weld zone (WZ) was predominantly composed of AF, and the heat-affected zone (HAZ) primarily consisted of TM and LM. Consequently, the welded joint still exhibited excellent mechanical properties, particularly toughness, despite higher welding heat input. The average tensile strength reached 928 MPa, and the impact absorbed energy at -40 °C for the WZ and HAZ were 54 J and 126 J, respectively. In addition, the application of triple-wire welding for ultrahigh-strength steel (UHSS) demonstrated a significant enhancement in post-weld deposition rate, with increases of 106% and 38% compared to single-wire and twin-wire welding techniques, respectively. This process not only utilized flux-cored wire to enhance the mechanical properties of joints but also achieved high deposition rate welding.

The Role of Interferon Regulatory Factors in Liver Diseases.

The interferon regulatory factors (IRFs) family comprises 11 members that are involved in various biological processes such as antiviral defense, cell proliferation regulation, differentiation, and apoptosis. Recent studies have highlighted the roles of IRF1-9 in a range of liver diseases, including hepatic ischemia-reperfusion injury (IRI), alcohol-induced liver injury, Con A-induced liver injury, nonalcoholic fatty liver disease (NAFLD), cirrhosis, and hepatocellular carcinoma (HCC). IRF1 is involved in the progression of hepatic IRI through signaling pathways such as PIAS1/NFATc1/HDAC1/IRF1/p38 MAPK and IRF1/JNK. The regulation of downstream IL-12, IL-15, p21, p38, HMGB1, JNK, Beclin1, ß-catenin, caspase 3, caspase 8, IFN-γ, IFN-ß and other genes are involved in the progression of hepatic IRI, and in the development of HCC through the regulation of PD-L1, IL-6, IL-8, CXCL1, CXCL10, and CXCR3. In addition, IRF3-PPP2R1B and IRF4-FSTL1-DIP2A/CD14 pathways are involved in the development of NAFLD. Other members of the IRF family also play moderately important functions in different liver diseases. Therefore, given the significance of IRFs in liver diseases and the lack of a comprehensive compilation of their molecular mechanisms in different liver diseases, this review is dedicated to exploring the molecular mechanisms of IRFs in various liver diseases.

Multi-FusNet: fusion mapping of features for fine-grained image retrieval networks.

As the diversity and volume of images continue to grow, the demand for efficient fine-grained image retrieval has surged across numerous fields. However, the current deep learning-based approaches to fine-grained image retrieval often concentrate solely on the top-layer features, neglecting the relevant information carried in the middle layer, even though these information contains more fine-grained identification content. Moreover, these methods typically employ a uniform weighting strategy during hash code mapping, risking the loss of critical region mapping-an irreversible detriment to fine-grained retrieval tasks. To address the above problems, we propose a novel method for fine-grained image retrieval that leverage feature fusion and hash mapping techniques. Our approach harnesses a multi-level feature cascade, emphasizing not just top-layer but also intermediate-layer image features, and integrates a feature fusion module at each level to enhance the extraction of discriminative information. In addition, we introduce an agent self-attention architecture, marking its first application in this context, which steers the model to prioritize on long-range features, further avoiding the loss of critical regions of the mapping. Finally, our proposed model significantly outperforms existing state-of-the-art, improving the retrieval accuracy by an average of 40% for the 12-bit dataset, 22% for the 24-bit dataset, 16% for the 32-bit dataset, and 11% for the 48-bit dataset across five publicly available fine-grained datasets. We also validate the generalization ability and performance stability of our proposed method by another five datasets and statistical significance tests. Our code can be downloaded from https://github.com/BJFU-CS2012/MuiltNet.git.

Development and validation of a nomogram for major adverse cardiovascular events after chronic total occlusion percutaneous coronary intervention for ischemic heart failure.

BACKGROUND: Chronic total occlusion percutaneous coronary intervention (CTO-PCI) is an available means of revascularization in patients with ischemic heart failure (IHF). However, the prognosis of IHF patients undergoing CTO-PCI remains unclear due to the lack of reliable clinical predictive tools. AIM: This study aimed to establish a nomogram for major adverse cardiovascular events (MACE) after CTO-PCI in IHF patients. METHODS: Sixty-seven potential predictive variables for MACE in 560 IHF patients undergoing CTO-PCI were screened using least absolute shrinkage and selection operator regression. A nomogram was constructed based on multivariable Cox regression to visualize the risk of MACE, and then evaluation was carried out using the concordance index (C-index), time-independent receiver operating characteristic (timeROC) curves, calibration curves, and decision curve analysis (DCA). RESULTS: During a median follow-up of 32.0 months, there were 208 MACE occurrences. Seven variables were selected for nomogram construction: age, left ventricular ejection fraction, left ventricular end-diastolic diameter, N-terminal precursor B-type diuretic peptide, bending, and use of intravascular ultrasound and beta-blockers. The C-index was 0.715 (0.680-0.750) and the internal validation result was 0.715 (0.676-0.748). The timeROC area under the curve at 6 months, 1 year, and 2 years was 0.750 (0.653-0.846), 0.747 (0.690-0.804), and 0.753 (0.708-0.798), respectively. The calibration curves and DCA showed the nomogram had acceptable calibration and clinical applicability. CONCLUSIONS: We developed a simple and efficient nomogram for MACE after CTO-PCI in IHF patients, which helps in early risk stratification and postoperative management optimization.

Genesis of the hydrothermal gold system in the kibaran metallogenic province (D.R.Congo): A review for the Twangiza-Namoya gold belt.

The Twangiza-Namoya Gold Deposit within the Kibaran Belt of the Democratic Republic of the Congo represents a crucial manifestation of the hydrothermal gold system. This review investigates its intricate origin and the subsequent metallogenic evolution that has shaped its present-day characteristics and offers a systematic categorization based on its deposition processes and geotectonic settings. The findings reveal that the gold deposits are predominantly derived from sedimentary fluid sources. Within this vast metallogenic province, two stages of gold deposition have been constrained: (a) the early-stage formation related to the accretion of Rodinia assembly with subduction-collisional event where diagenesis cemented the syngenetic pyrite carbonaceous sediments and (b) the later stage deposition related to the continent-collisional event during the last stage of Rodinia supercontinent amalgamation. Previous isotopic investigations, with a particular emphasis on pyrite sulfur isotopes on both host rocks and vein-bearing sulfides, have been instrumental in tracing the origins of gold-bearing fluids in the study region. The isotopic variance in the four deposits: Twangiza (-5.2 % to +3 %, avg. -0.3 %), Kamituga (-0.6 % to -0.9.1 %, avg. -5%), Lugushwa (+3.0 % to -18.4 %), and Namoya, on the southernmost end, has a vast range but with much heavier isotope compositions, ranging between +1.3 % and to +22.6 %, with an average of +12.2 %. The data predominantly points to the sedimentary origins of ore fluids in the Twangiza-Namoya Gold belt, highlighting the pivotal role of sedimentary processes in shaping the metallogenic landscape of the region. The fluids inclusions depicted the deposits to be formed from H2O-Nacl-H2O with abundant CH4 and N2 ore-forming fluid, moderate temperature (350-500 °C), and low salinity. The overall results confirm the genetic style of the Twangiza-Namoya Gold Belt to be an orogenic gold-style deposit that was emplaced during the early Neoproterozoic era in low greenschist facies terrain.

Tumor micro-environment induced TRAIL secretion from engineered macrophages for anti-tumor therapy.

The high plasticity and long-term persistency make macrophages excellent vehicles for delivering anti-tumor cytokines. Macrophage delivery of chemokines and cytokines shows potential in tumor therapy. TRAIL, a promising anti-tumor cytokine, induces apoptosis in tumor cells with low toxicity to normal cells. However, its off-target toxicity and limited stability have limited its clinical progress. Here, we engineered macrophages with Mono-TRAIL and Tri-TRAIL and found that Tri-TRAIL had higher cytotoxic activity against tumor cells than Mono-TRAIL in vitro. To target the tumor microenvironment (TME), we generated macrophages secreting trimeric TRAIL (Tri-TRAIL-iM) induced by the TME-specific promoter Arg1. The Tri-TRAIL-iM cells displayed high specific activatable activity in cell-based co-culture assay and tumor-baring mice models. In addition, we demonstrated that compared to macrophages over-expressing TRAIL under a non-inducible promoter, Tri-TRAIL-iM could more effectively induce apoptosis in cancer cells, inhibit tumor growth, and reduce systemic side effects. This strategy of inducing TRAIL delivery holds great potential for cancer therapy. It is promising to be combined with other engineering methods to maximize the therapeutic effects of solid tumors.

Uptake of zinc from the soil to the wheat grain: Nonlinear process prediction based on artificial neural network and geochemical data.

Trace elements in plants primarily derive from soils, subsequently influencing human health through the food chain. Therefore, it is essential to understand the relationship of trace elements between plants and soils. Since trace elements from soils absorbed by plants is a nonlinear process, traditional multiple linear regression (MLR) models failed to provide accurate predictions. Zinc (Zn) was chosen as the objective element in this case. Using soil geochemical data, artificial neural networks (ANN) were utilized to develop predictive models that accurately estimated Zn content within wheat grains. A total of 4036 topsoil samples and 73 paired rhizosphere soil-wheat samples were collected for the simulation study. Through Pearson correlation analysis, the total content of elements (TCEs) of Fe, Mn, Zn, and P, as well as the available content of elements (ACEs) of B, Mo, N, and Fe, were significantly correlated with the Zn bioaccumulation factor (BAF). Upon comparison, ANN models outperformed MLR models in terms of prediction accuracy. Notably, the predictive performance using ACEs as input factors was better than that using TCEs. To improve the accuracy, a two-step model was established through multiple testing. Firstly, ACEs in the soil were predicted using TCEs and properties of the rhizosphere soil as input factors. Secondly, the Zn BAF in grains was predicted using ACE as input factors. Consequently, the content of Zn in wheat grains corresponding to 4036 topsoil samples was predicted. Results showed that 85.69 % of the land was suitable for cultivating Zn-rich wheat. This finding offers a more accurate method to predict the uptake of trace elements from soils to grains, which helps to warn about abnormal levels in grains and prevent potential health risks.

Adan: Adaptive Nesterov Momentum Algorithm for Faster Optimizing Deep Models.

In deep learning, different kinds of deep networks typically need different optimizers, which have to be chosen after multiple trials, making the training process inefficient. To relieve this issue and consistently improve the model training speed across deep networks, we propose the ADAptive Nesterov momentum algorithm, Adan for short. Adan first reformulates the vanilla Nesterov acceleration to develop a new Nesterov momentum estimation (NME) method, which avoids the extra overhead of computing gradient at the extrapolation point. Then Adan adopts NME to estimate the gradient's first- and second-order moments in adaptive gradient algorithms for convergence acceleration. Besides, we prove that Adan finds an ϵ-approximate first-order stationary point within O(ϵ-3.5) stochastic gradient complexity on the non-convex stochastic problems (e.g.deep learning problems), matching the best-known lower bound. Extensive experimental results show that Adan consistently surpasses the corresponding SoTA optimizers on vision, language, and RL tasks and sets new SoTAs for many popular networks and frameworks, eg ResNet, ConvNext, ViT, Swin, MAE, DETR, GPT-2, Transformer-XL, and BERT. More surprisingly, Adan can use half of the training cost (epochs) of SoTA optimizers to achieve higher or comparable performance on ViT, GPT-2, MAE, etc, and also shows great tolerance to a large range of minibatch size, e.g.from 1k to 32k. Code is released at https://github.com/sail-sg/Adan, and has been used in multiple popular deep learning frameworks or projects.

The reduction of the carbon footprint of municipal solid waste management via source classification and supporting strategies: An analysis for the megacity of Shenzhen.

Municipal solid waste (MSW) management is a critical concern in megacities that depend heavily on external material and energy inputs but lack space for waste disposal. MSW treatment is a significant contributor to carbon emissions. The implementation of source classification improved the overall MSW management system and enhanced resource recovery from MSW. However, the precise contribution of source classification to carbon emissions reduction remains unclear. This study aimed to analyze the carbon emissions evolution in the MSW management of Shenzhen, a prototypical megacity in China, using data from 2006 to 2020 and employing carbon footprint assessment methodologies. The results demonstrated that source classification reduced the carbon emissions from 0.19-0.25 to 0.14-0.18 t CO2-eq/t MSW when considering the contribution of the urban environmental sanitation management department. The entire MSW management system becomes a carbon sink when considering recyclables collected by commercial enterprises. Although the source classification complicated the collection and transportation of MSW, the carbon offset effect of recycling food waste and recyclables was more significant than that of carbon emissions from collection and transport. Moreover, the landfill gas recovery rate critically influenced the carbon emissions of landfill-based MSW management systems. In contrast, the recovery of plastics was crucial for determining carbon emissions from incineration-based MSW management systems.

Lead Iodide Redistribution Enables In Situ Passivation for Blading Inverted Perovskite Solar Cells with 24.5% Efficiency.

Blade-coating stands out as an alternative for fabricating scalable perovskite solar cells. However, it demands special control of the precursor composition regarding nucleation and crystallization and currently exhibits lower performance than the spin-coating process. It is mainly the resulting film morphology and excess lead iodide (PbI2) distribution that influences the optoelectronic properties. Here, the effectiveness of introducing N-Methyl-2-pyrrolidone (NMP) to regulate the structure of the perovskite layer and the redistribution of PbI2 is found. The introduction of NMP leads to the accumulation of excess PbI2, mainly on the top surface, reducing residual PbI2 at the perovskite buried interface. This not only facilitates the passivation of perovskite grain boundaries but also eliminates the potential degradation of the PbI2 triggered by light illumination in the perovskite buried interface. The optimized NMP-modified inverted perovskite solar cell achieves a champion efficiency of 24.5%, among the highest reported blade-coated perovskite solar cells. Furthermore, 13.68 cm2 blading perovskite solar modules are fabricated and demonstrate an efficiency of up to 20.4%. These findings underscore that with proper modulation of precursor composition, blade-coating can be a feasible and superior alternative for manufacturing high-quality perovskite films, paving the way for their large-scale applications in photovoltaic technology.