Swin MoCo: Improving parotid gland MRI segmentation using contrastive learning.

BACKGROUND: Segmentation of the parotid glands and tumors by MR images is essential for treating parotid gland tumors. However, segmentation of the parotid glands is particularly challenging due to their variable shape and low contrast with surrounding structures. PURPOSE: The lack of large and well-annotated datasets limits the development of deep learning in medical images. As an unsupervised learning method, contrastive learning has seen rapid development in recent years. It can better use unlabeled images and is hopeful to improve parotid gland segmentation. METHODS: We propose Swin MoCo, a momentum contrastive learning network with Swin Transformer as its backbone. The ImageNet supervised model is used as the initial weights of Swin MoCo, thus improving the training effects on small medical image datasets. RESULTS: Swin MoCo trained with transfer learning improves parotid gland segmentation to 89.78% DSC, 85.18% mIoU, 3.60 HD, and 90.08% mAcc. On the Synapse multi-organ computed tomography (CT) dataset, using Swin MoCo as the pre-trained model of Swin-Unet yields 79.66% DSC and 12.73 HD, which outperforms the best result of Swin-Unet on the Synapse dataset. CONCLUSIONS: The above improvements require only 4 h of training on a single NVIDIA Tesla V100, which is computationally cheap. Swin MoCo provides new approaches to improve the performance of tasks on small datasets. The code is publicly available at https://github.com/Zian-Xu/Swin-MoCo.

Promoted Photocatalytic Hydrogen Evolution by Tuning the Electronic State of Copper Sites in Metal-Organic Supramolecular Assemblies.

The electronic state in terms of charge and spin of metal sites is fundamental to govern the catalytic activity of a photocatalyst. Herein, we show that modulation of the electronic states of Cu sites, without changing the coordination environments, of two metal-organic supramolecular assemblies based on π⋅⋅⋅π stacking can significantly improve photocatalytic activity. The use of these heterogeneous photocatalysts, without using noble metal cocatalysts, resulted in an increase of the hydrogen production rate from 522 to 3620 µmol h-1 g-1 . A systematical analysis revealed that the charge density and spin density of the metal centers are efficiently modulated via the modulation of the coordination fields around active copper (II) centers by the variation of the non-coordination groups of terminal ligands, leading to the significant enhancement of photocatalytic activity. This work provides an insight into the electronic state of active metal centers for designing high-performance photocatalysts.

A MOF-derived hierarchical CoP@ZnIn2S4 photocatalyst for visible light-driven hydrogen evolution.

A hierarchical CoP@ZnIn2S4 photocatalyst was prepared via a MOF-templated strategy. Owing to the unique composition and morphology that can facilitate the separation of photoexcited carriers, enhance light absorption and provide high surface area, CoP@ZnIn2S4 exhibited a H2 evolution rate of 0.103 mmol h-1 and remained stable over 24 hours.

Modulation of Z-Scheme Heterojunction Interface between Ultrathin C3N5 Nanosheets and Metal-Organic Framework for Boosting Photocatalysis.

Fabricating heterojunction photocatalysts for H2 production is promising for the development of clean energy. For boosting the photocatalytic activity, modulating the heterojunction interface can facilitate the electron-hole separation and solar energy utilization, but it is highly challenging in synthesis. In this work, by facilely exfoliating the bulk C3N5, ultrathin C3N5 nanosheets (N-CN) with large surface area, improved light absorption, and efficient charge transport were synthesized and further applied to the construction of NH2-UiO-66/N-CN heterojunctions. The optimized NH2-UiO-66/N-CN-2 exhibits high hydrogen evolution rate and cycling stability with Pt as the cocatalyst. Combined with the experimental results, the density functional theory calculation reveals that the high photocatalytic performance is attributed to the promoted photogenerated carrier transfer by the formation of well-contacted and stable Z-scheme heterojunction interface. This contribution renders an insight into the modulation of the heterojunction interface for enhancing the activity of MOF-based photocatalysts.

Bimetallic Cage-Based Metal-Organic Frameworks for Electrochemical Hydrogen Evolution Reaction with Enhanced Activity.

A cage-based metal-organic framework (Ni-NKU-101) with biphenyl-3,3',5,5'-tetracarboxylic acid was synthesized via solvothermal method. Ni-NKU-101 contains two types of cages based on trinuclear and octa-nuclear nickel-clusters that are connected with each other by the 4-connected ligands, to form a 3D framework with a new topology. A mixed-metal strategy was used to synthesize isostructural bimetallic MOFs of Mx Ni1-x -NKU-101 (M=Mn, Co, Cu, Zn). The electrocatalytic studies showed that the hydrogen evolution reaction (HER) activity of Cux Ni1-x -NKU-101 is much higher than that of other Mx Ni1-x -NKU-101 catalysts in acidic aqueous solution, owing to the synergistic effect of the bimetallic centers. The optimized Cu0.19 Ni0.81 -NKU-101 has an overpotential of 324 mV at 10 mA cm-2 and a Tafel slope of 131 mV dec-1 . The mechanism of HER activity over these bimetallic MOF-based electrocatalysts are discussed in detail.

Isolation of individual isoflavone species from soybean by solvent extraction followed by the combination of macroporous resin and aluminium oxide separation.

Growing interest in the health benefits of soy isoflavones has led to research in the isolation of individual isoflavone species for further application. Herein, we develop a new strategy to isolate daidzein, genistein, daidzin and genistin in soybean. We investigated the impact of solvents used and the extraction time on the extracted isoflavone contents from soybean. A 30-min extraction with 65% aqueous methanol gave a total isoflavone yield of 345 mg/100 g soybean, the highest value among tested conditions. Further, we proposed a two-stage adsorption/desorption chromatography comprising macroporous resin and aluminium oxide to isolate isoflavone. First, HP-20 resin was used to separate the glucosidic and aglyconic forms of isoflavone, then individual species of isoflavone could be isolated using aluminium oxide by specific retention of 5-hydroxy isoflavone. This process achieved overall high recovery (82-97%) and purity (92-95%) of the four isoflavones, which confirms a high separating efficiency for isoflavones from soybean.

An Efficient and Stable MoS2 /Zn0.5 Cd0.5 S Nanocatalyst for Photocatalytic Hydrogen Evolution.

Photocatalytic hydrogen evolution by water splitting is highly important for the application of hydrogen energy and the replacement of fossil fuel by solar energy, which needs the development of efficient catalysts with long-term catalytic stability under light irradiation in aqueous solution. Herein, Zn0.5 Cd0.5 S solid solution was synthesized by a metal-organic framework-templated strategy and then loaded with MoS2 by a hydrothermal method to fabricate a MoS2 /Zn0.5 Cd0.5 S heterojunction for photocatalytic hydrogen evolution. The composition of MoS2 /Zn0.5 Cd0.5 S was fine-tuned to obtain the optimized 5 wt % MoS2 /Zn0.5 Cd0.5 S heterojunction, which showed a superior hydrogen evolution rate of 23.80 mmol h-1 g-1 and steady photocatalytic stability over 25 h. The photocatalytic performance is due to the appropriate composition and the formation of an intimate interface between MoS2 and Zn0.5 Cd0.5 S, which endows the photocatalyst with high light-harvesting ability and effective separation of photogenerated carriers.

A Metal-Organic-Framework-Derived (Zn0.95Cu0.05)0.6Cd0.4S Solid Solution as Efficient Photocatalyst for Hydrogen Evolution Reaction.

Photocatalytic water splitting taking the advantage of using solar energy directly is one of the most effective strategies for hydrogen evolution. The development of facile methods for synthesizing highly efficient and stable photocatalysts for hydrogen production still remains a great challenge. Herein, a metal-organic framework (MOF)-templated strategy was designed for the synthesis of solid solutions of (Zn0.95Cu0.05)1-xCdxS that exhibit outstanding photocatalytic hydrogen production reaction activity. More importantly, efficient light capturing ability and photogenerated charges separation were accomplished via fine-tuning the composition of the photocatalysts by adjusting the concentrations of doping metals in the template MOFs. Under visible light (λ > 420 nm), an optimized nanocatalyst, (Zn0.95Cu0.05)0.6Cd0.4S, exhibited a higher durability and satisfied photocatalytic hydrogen evolution rate of 4150.1 µmol g-1 h-1 of water splitting.

Phosphorylation of Isoflavones by Bacillus subtilis BCRC 80517 May Represent Xenobiotic Metabolism.

The soy isoflavones daidzein (DAI) and genistein (GEN) have beneficial effects on human health. However, their oral bioavailability is hampered by their low aqueous solubility. Our previous study revealed two water-soluble phosphorylated conjugates of isoflavones, daidzein 7-O-phosphate and genistein 7-O-phosphate, generated via biotransformation by Bacillus subtilis BCRC80517 cultivated with isoflavones. In this study, two novel derivatives of isoflavones, daidzein 4'-O-phosphate and genistein 4'-O-phosphate, were identified by HPLC-ESI-MS/MS and 1H, 13C, and 31P NMR, and their biotransformation roadmaps were proposed. Primarily, isoflavone glucosides were deglycosylated and then phosphorylated predominantly into 7-O-phosphate conjugates with traces of 4'-O-phosphate conjugates. Inevitably, trace quantities of glucosides were converted into 6″-O-succinyl glucosides. GEN was more efficiently phosphorylated than DAI. Nevertheless, the presence of GEN prolonged the time until the exponential phase of cell growth, whereas the other isoflavones showed little effect on cell growth. Our findings provide new insights into the novel microbial phosphorylation of isoflavones involved in xenobiotic metabolism.

Home-based mobile cardio-pulmonary rehabilitation consultant system.

Cardiovascular diseases are the most popular cause of death in the world recently. For postoperatives, cardiac rehabilitation is still asked to maintain at home (phase II) to improve cardiac function. However, only one third of outpatients do the exercise regularly, reflecting the difficulty for home-based healthcare: lacking of monitoring and motivation. Hence, a cardio-pulmonary rehabilitation system was proposed in this research to improve rehabilitation efficiency for better prognosis. The proposed system was built on mobile phone and receiving electrocardiograph (ECG) signal from a wireless ECG holter via Bluetooth connection. Apart from heart rate (HR) monitor, an ECG derived respiration (EDR) technique is also included to provide respiration rate (RR). Both HR and RR are the most important vital signs during exercise but only used one physiological signal recorder in this system. In clinical test, there were 15 subjects affording Bruce Task (treadmill) to simulate rehabilitation procedure. Correlation between this system and commercial product (Custo-Med) was up to 98% in HR and 81% in RR. Considering the prevention of sudden heart attack, an arrhythmia detection expert system and healthcare server at the backend were also integrated to this system for comprehensive cardio-pulmonary monitoring whenever and wherever doing the exercise.