深度学习在医学影像学领域应用研究进展

阐述深度学习方法、原理和主要模型结构,以脑部疾病和乳腺癌为例分析深度学习在国内医学影像学领域的应用情况,总结其局限性,包括图像获取难度高、数据缺乏完整性、数据采集标准化不足及图像像素分辨率不高等方面。     

膝关节半月板热身运动前后磁共振T2时间变化研究

目的 通过比较分析热身运动前后膝关节半月板磁共振T₂时间的变化,探讨利用T₂时间变化反映热身运动前后半月板变化的可行性。方法 选择40例无膝关节相关症状的志愿者,其中男性15例,女性25例;患者年龄22~30岁,平均年龄26.32±2.1岁。在同等强度的热身运动前后对每位志愿者行右膝关节的MR T₂ mapping序列成像,分别测量内、外侧半月板前角、后角、体部的T₂时间,采用配对t检验比较热身运动前后膝关节半月板T₂时间变化。结果 热身运动后膝关节半月板各个部位的T₂值均增高（P=0.000）;外侧半月板体部T₂值增高幅度小于内侧半月板前角（P<0.05）、内侧半月板后角（P<0.05）及外侧半月板前角（P<0.05）。结论 热身运动可以增加膝关节半月板T₂时间,提示可以增加半月板内含水量,改善半月板的生理活性,对半月板起到一定的保护作用;MR T₂ mapping成像技术可以评价运动前后半月板T₂时间,是评估半月板的1项可靠手段。     

Optimisation of protein microarray techniques for analysis of the plasma proteome: Minimisation of non-specific binding interactions

Components of the plasma proteome, particularly serum albumin, have been shown to compromise the accuracy of protein microarray technologies through non-specific binding interactions. Optimisation of array conditions is imperative to help address these problems. Here we demonstrate how modifications to array printing conditions and processing methodology can influence the reliability of data output. In particular, we demonstrate that whilst some glycerol is necessary to maintain specific binding signals, it also increases non-specific binding of albumin. Concentrations of 20% glycerol in the printing buffers are therefore recommended. The findings presented here provide opportunities for increased accuracy in plasma protein detection for possible future diagnostic applications.     

Spatially and temporally controlled immune cell interactions using microscale tools

1. Download : Download high-res image (234KB)
2. Download : Download full-size image

     

Illustration of charge transfer in graphene-coated hexagonal ZnO photocatalysts using Kelvin probe force microscopy

A graphene coated hexagonal ZnO (HZO@Gr) with enhanced activity in photocatalysis was synthesized. However, the photoinduced charge transfer behavior and the beneficial role of graphene in promoting photocatalytic reactions have not been sufficiently investigated experimentally. In this paper, the surface potentials of the ±(0001)-polar plane of HZO (Zn-polar plane and O-polar plane), graphene, graphene/Zn-polar plane and graphene/O-polar plane were measured using Kelvin probe force microscopy (KPFM). On the basis of the KPFM results, the respective Fermi levels were calculated and the internal electric field (IEF) of HZO was confirmed. Taking the IEF of HZO into consideration, the three-dimensional band diagrams of the HZO@Gr composites in methyl blue (MB) solution in the dark and under UV-visible irradiation after equilibrium were proposed. Accordingly, it is found that there could emerge different interactions between graphene and HZO at the ±(0001)-polar plane of HZO. Furthermore, the photogenerated holes and electrons tend to migrate to opposite directions. With the participation of graphene and IEF, the composites show a decrease in possibility of charge recombination. As a result, the active groups, namely ˙OH and ˙O₂⁻ radicals, could be mainly generated at/near the O-polar plane and Zn-polar plane, respectively. This work can serve as a supplemental explanation of the charge transfer during the photocatalytic process at the polar ZnO/graphene composite surface.

The Fermi levels and three-dimensional band diagrams of the synthesized HZO@Gr composites in methyl blue (MB) solution before and after equilibrium were assumed.     

Electrooxidation of Pd–Cu NP loaded porous carbon derived from a Cu-MOF

The development of new non-platinum catalysts for alcohol electrooxidation is of utmost importance. In this work, a bimetallic Pd–Cu loaded porous carbon material was first synthesized from a Cu-based metal–organic framework (MOF). The Cu loaded porous carbon was pre-synthesized through calcinating the Cu-based MOF under a N₂ atmosphere. After loading Pd onto the precursor and heating, Pd–Cu loaded porous carbon (Pd–Cu/C) was obtained for alcohol electrooxidation. Electrooxidation experiments revealed that this Pd–Cu bimetal loaded porous carbon assisted steady state electrolysis for alcohol oxidation in alkaline media. Moreover, different alcohols were electrooxidated using the present electrocatalyst for the purposes of discussing the oxidation mechanism. This electrooxidation study of Pd–Cu/C derived from a MOF demonstrates a good understanding of the electrooxidation of different alcohols, and provides useful guidance for developing new electrocatalyst materials for energy conversion and electronic devices.

We have synthesized Pd–Cu NP loaded porous carbon through the direct carbonization of a porous Cu based MOF for efficient electrooxidation.

There is an immediate need to develop direct alcohol fuel cells (DAFCs), which have been proven to be a fine source of energy, which could probably replace fossil fuels use to fulfil global energy demand. As one of the most significant electrocatalytic procedures, the electrooxidation of alcohols is an important process in DAFCs, and has gathered much attention and is attractive, due to high power density output and low pollutant emissions. Generally, Pt based materials are the most common electrocatalysts for alcohol electrooxidation reactions. However, the high cost and limited supply of Pt severely restricts its commercial application. Therefore, the development of new efficient and inexpensive non-platinum alternative materials to Pt-based catalysts is of utmost importance.Metal–organic frameworks (MOF) are assembled from metal ions linked by organic ligands, and are used in catalysis, guest molecule storage/separation, fluorescence, sensors and other devices.^1–3 Due to their highly ordered porous structures and large surface areas, MOFs can also be used as templates/precursors for preparing porous carbon materials through thermal treatment.^4–8 Several MOF derived carbon materials with good electrical conductivity are reported to show effective electrocatalytic performance,⁹ such as in the oxygen evolution reaction (OER),¹⁰ hydrogen evolution reaction (HER),¹¹ and oxygen reduction reaction (ORR).¹² Recently, a zeolitic imidazolate framework ZIF-8 was calcinated in order to prepare porous carbon with both micro- and meso-pores to support Pd electrocatalysts for methanol electrooxidation.¹³ Unfortunately, this could not efficiently limit the use of the noble metal, which gives challenges to scientists for further exploration.It has been demonstrated that the alloying of noble metals with transition metals has been used for the enhancement of catalytic activity and reduction of cost,¹⁴ because of the low cost and relatively high abundance of transition metals. The use of alloyed metal, Pd–M (where M is Cu, Co or Ni), binary electrocatalysts has been reported for effectively improving the catalytic properties.^15–17 For this purpose, introducing a non-noble metal into a noble metal bimetallic system will fulfil the demand for a new catalyst and become an area of interest nowadays. The second metal (such as Cu) will behave as a donor, while Pd has an empty d orbital to accept electrons, and it is assumed that its electronic properties will be more similar to Pt.¹⁸ Therefore, Pd–Cu bimetal loaded porous carbon derived from a Cu-MOF can provide a good electrocatalyst for alcohol oxidation.In this work, as shown in Scheme 1, we have synthesized Cu loaded porous carbon through the direct carbonization of a porous Cu based MOF. After loading Pd onto the precursor and heating, Pd–Cu bimetal loaded porous carbon (Pd–Cu/C) was obtained. Chronoamperometric studies revealed that this Pd–Cu bimetal loaded porous carbon assisted steady state electrolysis for alcohol oxidation in alkaline media. In addition, alcohols with different numbers of carbon atoms (such as ethanol, 1-propanol and 2-propanol) were also investigated for electrooxidation. It is important to note that the effectiveness of this bimetallic NP loaded carbon means that it can serve as a catalyst for the electrooxidation of low-molecular weight alcohols, which probably can be used as energy sources in portable electronic devices.Open in a separate windowScheme 1The preparation procedure for Pd–Cu/C derived from HKUST-1 and PdCl₂: (a) MOF HKUST-1; (b) Cu/C calcinated from HKUST-1; (c) PdCl₂ loaded on Cu/C; and (d) the Pd–Cu/C material.Here, a 3-D MOF, HKUST-1 (also called Cu₃(BTC)₂, BTC = 1,3,5-benzenetricarboxylate), was chosen as the precursor for preparing the Cu/C material, due to the structure having high porosity and it being a rich Cu source. The as-synthesized HKUST-1 was calcinated at 700 °C for 5 h under a N₂ atmosphere, and the Cu/C material was obtained. In addition, the guest species PdCl₂ was loaded onto the calcinated HKUST-1 through immersing the pre-calcinated HKUST-1 into a PdCl₂ ethanolic solution (1 mM) for 2 h (Scheme 1). The PdCl₂ loaded Cu/C (PdCl₂@Cu/C) was heated at 300 °C for 1 h under a N₂ atmosphere. Finally, an alloy of Pd and Cu loaded porous carbon material (Pd–Cu/C) was obtained and characterized through powder XRD, BET and XPS analyses.The PXRD data (Fig. 1a) from as-synthesized HKUST-1 powder and the bimetallic Pd–Cu NP loaded carbon porous material derived from HKUST-1 show that the samples contain bimetallic palladium and copper mostly. The XRD peak appearing at 43.3° corresponds to the (fcc) (111) facet plane of Cu. Due to Pd being dispersed homogenously at a low concentration through the sample, the XRD pattern could not display the obvious peak from Pd. However, inductively coupled plasma emission spectroscopy (ICP) data (Table S1^†) from the sample showed 0.76% Pd and 36.68% Cu, indicating the existence of Pd and Cu. The porosity of Pd–Cu/C was demonstrated through BET data, which shows N₂ adsorption of ∼150 cm³ g⁻¹. The Pd XPS spectrum showed two definite peaks at 335.5 and 341 eV, respectively assigned to 3d_5/2 and 3d_3/2 and matching well with Pd⁰. XPS peaks at 932.4 and 952.1 eV indicate the valence states of Cu ions in the Cu 2p_3/2 and Cu 2p_1/2 orbitals in the Pd–Cu/C material. Cu²⁺ is present in the porous carbon material, with respective peaks at 933.7 eV and 934.4 eV from CuO and Cu(OH)₂, with a prominent satellite observed in the 938–946 eV range. A few Pd²⁺ ions also exist in the sample due the easy oxidation of the surface. The Raman spectrum of Pd–Cu/C (Fig. S4^†) shows typical graphitic carbon. The results of the characterization studies clearly reveal that the nanoparticles have a Pd and Cu bimetallic nature.Open in a separate windowFig. 1(a) XRD data from HKUST-1 and Pd–Cu/C; (b) N₂ sorption isotherms for Pd–Cu/C; and XPS data from (c) Pd and (d) Cu in a sample of Pd–Cu/C.SEM images with EDS (Fig. 2a and b) results show that the sample contained much more copper than palladium, which clearly suggests that the presence of copper in the sample would probably be the reason for the expected electrooxidation of alcohols. It could be possible to replace the use of high-cost Pd or Pt based catalysts. The morphology of the Pd–Cu NPs was further characterized via TEM imaging and TEM element mapping (Fig. 2c, d and S5^†), demonstrating that the nano-sized NPs were dispersed homogeneously. The HR-TEM image in Fig. 2c gives insight into the bimetallic nature of the synthesized nanoparticles, with two noticeable lattice fringes (0.225 nm for Pd(111) and 0.202 nm for Cu(111)). The mean size of the Pd–Cu NPs was 7.38 nm, as shown in Fig. 2d. The homogenous distribution, with well-defined bimetallic Pd–Cu based carbon material, was good for the electrooxidation of alcohols. The electrochemical active surface area (ECSA) for Pd–Cu/C was high compared with commercial Pd/C, which suggested that the synthesized Pd–Cu/C has ample available surface area, mainly because of synergistic effects from the Cu-MOF based carbon material and the morphology of the electrocatalyst.Open in a separate windowFig. 2(a) SEM image of and (b) EDS data from Pd–Cu/C; (c) a TEM image of Pd–Cu NPs in the hybrid carbon material; and (d) the size distribution of the Pd–Cu NPs.In Fig. 3, CV profiles for commercial Pd/C and the presented Pd–Cu/C show two distinct peaks (forward (iF) and backward (iB) peaks) during the oxidation of methanol-containing 1 M KOH solution. The peak at −0.37 V indicates the oxidation of aforementioned carbonaceous species, such as Pd–CO_ads, along with newly formed alcohol adsorbates, following the removal of surface intermediates at lower potentials.¹⁹ For the forward peak potential, a shift in the iF value is observed, mainly because of Cu existing with Pd in the material. This results in the oxidation of poisonous species, such as Pd adsorbed CO, at higher potentials,²⁰ leading to such high activity. ATR-IR (Fig. S7^†) and GC analyses (Fig. S8^†) show the methanol oxidation reaction (MOR) pathway during the formation of the final CO₂ product. The catalytic activity of Pd–Cu/C is found to be ∼13 times higher than commercial Pd/C for methanol oxidation, demonstrating that the presence of Cu with Pd in Pd–Cu based catalysts increases CO oxidation because of a strong binding ability. Cu binds to CO more strongly than Pd, as a result of electronic structure differences,^21,22 thus preventing the electrode from undergoing CO poisoning, a major issue for Pd-based catalysts during the methanol oxidation reaction (MOR). The mechanism of methanol oxidation is shown in the ESI (eqn (8)–(10)).^† The high iF value for Pd–Cu/C can be ascribed to the fast formation of reactive intermediates, such as Pd–CH₂OH, Pd–COOH, Pd–H, Pd–(CHO)_ads, and Pd–(COOH)_ads.^23–25 The removal of these intermediates is necessary for a high current density. Furthermore, formaldehyde (HCHO), formic acid (HCOOH) and CO₂ would be the final products in the MOR.^26,27 Pd–Cu/C has good catalytic activity for the MOR, leading to further investigation into the electrooxidation of different alcohols, such as ethanol, 1-propanol, and 2-propanol (Fig. 4 and Table S2^†).Open in a separate windowFig. 3CV curves from Pd/C and Pd–Cu/C electrocatalysts during CH₃OH (1 M) oxidation in 1 M KOH solution, at a scan rate of 50 mV s⁻¹, at room temperature.Open in a separate windowFig. 4(a) CV curves from: the Pd–Cu/C electrocatalyst for C₁–C₃ aliphatic alcohol (1 M) oxidation in KOH (1 M) solution; and (b) Pd–Cu/C in 1 M EtOH, 1-propanol and 2-propanol at a scan rate of 50 mV s⁻¹ at room temperature.The current densities for different alcohol oxidation processes are summarized in Fig. 4a and b. The normalized iF (calculated using Pd mass) for the MOR (∼4643 mA mg⁻¹) was higher than for three other alcohols, i.e., it was ∼139, ∼94 and ∼26.5 mA mg⁻¹ for ethanol, 1-propanol and 2-propanol, respectively. The iF/iB ratio for methanol is ∼12 times higher than that for ethanol, ∼13 times that for 1-propanol and ∼4 times that for 2-propanol. The reactivity order for Pd–Cu/C is methanol > ethanol > 1-propanol > 2-propanol. 2-Propanol electrooxidation showed a lower current density on a Pd–Cu/C electrode in alkaline medium, although iF/iB is ∼5.4. The negative shift in the onset of the ethanol oxidation reaction (EOR) suggested that a high copper content with very low amount of Pd was suitable for EOR kinetics using a Pd–Cu/C catalyst. The ethoxy (CH₃CO)_ads was strongly adsorbed, and blocked hydrogen absorption/adsorption. The current intensity of iF increased due to the formation of fresh Pd–OH, through stripping carbonaceous residue from the electrode (eqn (11)–(14)^†). In addition, the increased current at high potentials sharply reached the largest value then started to decline, because a PdO layer formed on the electrode, blocking the further adsorption of reactive species.²⁸ ATR-IR spectra (Fig. S9^†) show the presence of CO₂ and CO_ads, whereas bands appear at 1670 cm⁻¹ and 1390 cm⁻¹ because of the formation of acetic acid.²⁹ The Pd–Cu/C electrocatalyst has the potential to oxidize the intermediate to the final product, CO₂, during EOR to some extent; Cu promotes oxidation through increasing the production of OH_ads⁻/H₂O to eliminate the intermediate CH₃CO_ads simultaneously on Pd.³⁰ The stability of Pd–Cu/C in all four alcohols (methanol, ethanol, 1-propanol and 2-propanol) was studied using chronoamperometry at a potential of −0.25 V, as shown in Fig. S2.^† The slow current decay showed that the stability of the Pd–Cu/C electrocatalyst in methanol is best. Comparing the results, the current for methanol oxidation was higher than that for the other three alcohols. However, the oxidation currents from ethanol and 1-propanol were larger than that from 2-propanol. This suggested that Pd–Cu/C is less stable and shows lower anti-poisoning ability during 2-propanol oxidation in an alkaline medium.During this oxidation, 1-propanol oxidizes to propanal first, and its further oxidation results in the formation of a stable product, propanoic acid (Scheme S1^†). 1-Propanol is converted, with its carboxylate as the major product, as verified using ATR-IR (Fig. S10^†). 2-Propanol forms acetone as an intermediate product, leading to the poisoning of the electrode.³¹ ATR-IR spectra (Fig. S11^†) of Pd–Cu/C also confirm that the electrocatalyst follows a dual pathway through acetone and propene intermediates to oxidize to CO₂ finally (Scheme S2^†).^32–34 However, acetone formation is kinetically favored.³⁵ The results show that the location of the –OH group in the alcohol influences the electrooxidation reaction kinetics. In contrast methanol oxidation using Pd–Cu/C has much higher catalytic activity than ethanol, 1-propanol and 2-propanol oxidation, which makes it a good candidate for direct methanol fuel cells.In summary, we have first synthesized a bimetallic Pd–Cu NP loaded porous carbon material from a Cu-based MOF for alcohol electrooxidation. The Cu loaded porous carbon was pre-synthesized by calcinating the Cu-based MOF HKUST-1 under a N₂ atmosphere. Afterwards, Pd–Cu NP loaded porous carbon was obtained for alcohol electrooxidation. Electrooxidation experiments revealed that Pd–Cu/C was suitable for steady state electrolysis for alcohol oxidation in alkaline media. In addition, different alcohols were electrooxidated using the present electrocatalyst to discuss the oxidation mechanism. This electrooxidation study of Pd–Cu/C derived from a MOF offers good understanding into the electrooxidation of different alcohols and it could provide useful guidance for the development of new electrocatalyst materials.     

Efficient and reusable SBA-15-immobilized Brønsted acidic ionic liquid for the ketalization of cyclohexanone with glycol

Ketalization of cyclohexanone with glycol has been carried out using molecular sieve SBA-15 immobilized Brønsted acidic ionic liquid catalyst. The properties of the heterogeneous catalysts were characterized by elemental analysis, Fourier transform infrared (FT-IR) spectra, scanning electron microscopy (SEM), thermogravimetry/differential scanning calorimetry (TG/DSC), and N₂ adsorption–desorption (BET). The results suggested that Brønsted acidic ionic liquid [BSmim][HSO₄] had been successfully immobilized on the surface of SBA-15 and the catalytic performance evaluation demonstrated that the catalyst BAIL@SBA-15 exhibited excellent catalytic activities in the ketalization of cyclohexanone with glycol. In addition, the effects of reaction temperature, catalyst loading, reaction time, and reactant molar ratio have also been investigated in detail, and a general reaction mechanism for the ketalization of cyclohexanone with glycol was given. The SBA-15 immobilized ionic liquid can be recovered easily and after reusing for 5 times in the ketalization reaction, the catalyst could still give satisfactory catalytic activity.

Molecular sieve SBA-15 immobilized Brønsted acidic ionic liquid: an efficient and recyclable catalyst for the ketalization of cyclohexanone with glycol.     

Synthesis of a novel CO2-based alcohol amine compound and its usage in obtaining a water- and solvent-resistant coating

A five-membered cyclo-carbonate, prepared by cycloaddition reaction from CO₂ and 1,4-butanediol diglycidyl ether, was reacted with excessive diamine and formed a urethane group-containing new product. Structural characterization was performed for the new alcohol amine, which can be applied to the manufacture of polyurethane coatings as a chain extender. The new chain extender-based polyurethane coatings exhibited excellent water, salt, and solvent resistance and promising mechanical strength. Importantly, the unique performance of the prepared polyurethane coatings should be ascribed to the introduction of a hydroxyl group in the polyurethane molecule. The strengthened hydrogen bonding enlarged the cohesion of the polyurethane coatings and prohibited the solvents from permeating.

The novelty of the work: a water- and solvent-resistant coating was produced from a novel CO₂-based alcohol amine chain extender.