[Retracted] Molecular hydrogen improves type 2 diabetes through inhibiting oxidative stress.

[This retracts the article DOI: 10.3892/etm.2020.8708.].

Polymeric Electronic Shielding Layer Enabling Superior Dendrite Suppression for All-Solid-State Lithium Batteries.

LiBH4 is one of the most promising candidates for use in all-solid-state lithium batteries. However, the main challenges of LiBH4 are the poor Li-ion conductivity at room temperature, excessive dendrite formation, and the narrow voltage window, which hamper practical application. Herein, we fabricate a flexible polymeric electronic shielding layer on the particle surfaces of LiBH4. The electronic conductivity of the primary LiBH4 is reduced by 2 orders of magnitude, to 1.15 × 10-9 S cm-1 at 25 °C, due to the high electron affinity of the electronic shielding layer; this localizes the electrons around the BH4- anions, which eliminates electronic leakage from the anionic framework and leads to a 68-fold higher critical electrical bias for dendrite growth on the particle surfaces. Contrary to the previously reported work, the shielding layer also ensures fast Li-ion conduction due to the fast-rotational dynamics of the BH4- species and the high Li-ion (carrier) concentration on the particle surfaces. In addition, the flexibility of the layer guarantees its structural integrity during Li plating and stripping. Therefore, our LiBH4-based solid-state electrolyte exhibits a high critical current density (11.43 mA cm-2) and long cycling stability of 5000 h (5.70 mA cm-2) at 25 °C. More importantly, the electrolyte had a wide operational temperature window (-30-150 °C). We believe that our findings provide a perspective with which to avoid dendrite formation in hydride solid-state electrolytes and provide high-performance all-solid-state lithium batteries.

Positive-Negative Receptive Field Reasoning for Omni-Supervised 3D Segmentation.

Hidden features in the neural networks usually fail to learn informative representation for 3D segmentation as supervisions are only given on output prediction, while this can be solved by omni-scale supervision on intermediate layers. In this paper, we bring the first omni-scale supervision method to 3D segmentation via the proposed gradual Receptive Field Component Reasoning (RFCR), where target Receptive Field Component Codes (RFCCs) is designed to record categories within receptive fields for hidden units in the encoder. Then, target RFCCs will supervise the decoder to gradually infer the RFCCs in a coarse-to-fine categories reasoning manner, and finally obtain the semantic labels. To purchase more supervisions, we also propose an RFCR-NL model with complementary negative codes (i.e., Negative RFCCs, NRFCCs) with negative learning. Because many hidden features are inactive with tiny magnitudes and make minor contributions to RFCC prediction, we propose Feature Densification with a centrifugal potential to obtain more unambiguous features, and it is in effect equivalent to entropy regularization over features. More active features can unleash the potential of omni-supervision method. We embed our method into three prevailing backbones, which are significantly improved in all three datasets on both fully and weakly supervised segmentation tasks and achieve competitive performances.

In-Situ-Generated Electron-Blocking LiH Enabling an Unprecedented Critical Current Density of Over 15 mA cm-2 for Solid-State Hydride Electrolytes.

LiBH4 is a promising solid-state electrolyte (SE) due to its thermodynamic stability to Li. However, poor Li-ion conductivities at room temperature, low oxidative stabilities, and severe dendrite growth hamper its application. In this work, a partial dehydrogenation strategy is adopted to in situ generate an electronic blocking layer dispersed of LiH, addressing the above three issues simultaneously. The electrically insulated LiH reduces the electronic conductivity by two orders of magnitude, leading to a 32.0-times higher critical electrical bias for dendrite growth on the particle surfaces than that of the counterpart. Additionally, this layer not only promotes the Li-ion conductance by stimulating coordinated rotations of BH4 - and B12 H12 2- , contributing to a Li-ion conductivity of 1.38 × 10-3 S cm-1 at 25 °C, but also greatly enhances oxidation stability by localizing the electron density on BH4 - , extending its voltage window to 6.0 V. Consequently, this electrolyte exhibits an unprecedented critical current density (CCD) of 15.12 mA cm-2 at 25 °C, long-term Li plating and stripping stability for 2700 h, and a wide temperature window for dendrite inhibition from -30 to 150 °C. Its Li-LiCoO2 cell displays high reversibility within 3.0-5.0 V. It is believed that this work provides a clear direction for solid-state hydride electrolytes.

DGRUnit: Dual graph reasoning unit for brain tumor segmentation.

Many fully automatic segmentation models have been created to solve the difficulty of brain tumor segmentation, thanks to the rapid growth of deep learning. However, few approaches focus on the long-range relationships and contextual interdependence in multimodal Magnetic Resonance (MR) images. In this paper, we propose a novel approach for brain tumor segmentation called the dual graph reasoning unit (DGRUnit). Two parallel graph reasoning modules are included in our proposed method: a spatial reasoning module and a channel reasoning module. The spatial reasoning module models the long-range spatial dependencies between distinct regions in an image using a graph convolutional network (GCN). The channel reasoning module uses a graph attention network (GAT) to model the rich contextual interdependencies between different channels with similar semantic representations. Our experimental results clearly demonstrate the superior performance of the proposed DGRUnit. The ablation study shows the flexibility and generalizability of our model, which can be easily integrated into a wide range of neural networks and further improve them. When compared to several state-of-the-art methods, experimental results show that the proposed approach significantly improves both visual inspection and quantitative metrics for brain tumor segmentation tasks.

Molecular hydrogen improves type 2 diabetes through inhibiting oxidative stress.

The aim of the present study was to investigate the potential therapeutic effects of molecular hydrogen on type 2 diabetes mellitus (T2DM) in rats. Following maintenance on a high-fat diet for 4 weeks, a T2DM model was established using an injection of 30 mg/kg streptozotocin via the caudal vein into Sprague-Dawley rats. On day 0 and Day 80, the blood samples were obtained from each rat for the measurement of biochemical indicators including blood lipids, fasting blood glucose, hepatic glycogen, fasting serum insulin, insulin sensitivity index, insulin resistance index, serum superoxide dismutase (SOD) and serum malondialdehyde (MDA) using an automatic biochemical analyzer. The kidneys and pancreas tissues were harvested for HE staining and Western blot assay of toll-like receptor 4 (TLR4), myeloid differentiation primary response 88 (MyD88), phosphorylated (p)-p65, p65, p-IκB and IκB. The results showed that in rats with T2DM, molecular hydrogen treatment decreased fasting blood glucose levels, increased hepatic glycogen synthesis and improved insulin sensitivity. Treatment with molecular hydrogen also increased the production of SOD whilst decreasing the production of MDA. In addition, molecular hydrogen alleviated the pathological changes exhibited by pancreatic islets and kidney during T2DM. Mechanistically, molecular hydrogen decreased TLR4 and MyD88 expression levels whilst also decreasing p65 and NF-κB inhibitor phosphorylation. In conclusion, molecular hydrogen exerted therapeutic effects against T2DM by improving hyperglycemia and inhibiting oxidative stress through mechanisms that are associated with the TLR4/MyD88/NF-κB signaling pathway.