Search

Objective. Optical coherence tomography (OCT) has become an essential imaging modality for the assessment of ophthalmic diseases. However, speckle noise in OCT images obscures subtle but important morphological details and hampers its clinical applications. In this work, a novel multi-task generative adversarial network (MGAN) is proposed for retinal OCT image denoising. Approach. To strengthen the preservation of retinal structural information in the OCT denoising procedure, the proposed MGAN integrates adversarial learning and multi-task learning. Specifically, the generator of MGAN simultaneously undertakes two tasks, including the denoising task and the segmentation task. The segmentation task aims at the generation of the retinal segmentation map, which can guide the denoising task to focus on the retina-related region based on the retina-attention module. In doing so, the denoising task can enhance the attention to the retinal region and subsequently protect the structural detail based on the supervision of the structural similarity index measure loss. Main results. The proposed MGAN was evaluated and analyzed on three public OCT datasets. The qualitative and quantitative comparisons show that the MGAN method can achieve higher image quality, and is more effective in both speckle noise reduction and structural information preservation than previous denoising methods. Significance. We have presented a MGAN for retinal OCT image denoising. The proposed method provides an effective way to strengthen the preservation of structural information while suppressing speckle noise, and can promote the OCT applications in the clinical observation and diagnosis of retinopathy.

Due to its unique properties, FeSe1−xSx superconductors have been prime candidates to investigate the unconventional superconducting mechanism in iron chalcogenides. In this study, we report the successful preparation of high-quality epitaxial FeSe1−xSx films with x ≤ 0.4 on different substrates via pulsed laser deposition. With different S content, it is found that CaF2 (100) single crystal is the most preferable substrate to use to grow superconducting FeSe1−xSx films. With the increment of S content, the nematic transition temperature (T s) of FeSe1−xSx films decreases simultaneously and disappears at x ≈ 0.2. In contrast to FeSe1−xTex films, the superconducting transition temperature (T c) of FeSe1−xSx films decreases continuously with S doping. And the T c shows no enhancement after the disappearance of T s, suggesting that the nematic transition does not positively impact the superconductivity in these films. In addition to the influence of different S content, it is found that the non-negligible in-plane tensile strain originating mostly from the mismatch between the FeSe1−xSx films and the substrate is the key factor suppressing the superconductivity. Our work can provide constructive guidance for preparing superconducting FeSe1−xSx films and gives a further understanding of the interplay of T s, strain, and superconductivity in iron-chalcogenide films.

Multi-modality examinations have been extensively applied in current clinical cancer management. Leveraging multi-modality medical images can be highly beneficial for automated tumor segmentation as they provide complementary information that could make the segmentation of tumors more accurate. This paper investigates CNN-based methods for automated nasopharyngeal carcinoma (NPC) segmentation using computed tomography (CT) and magnetic resonance (MR) images. Specially, a multi-modality convolutional neural network (M-CNN) is designed to jointly learn a multi-modal similarity metric and segmentation of paired CT-MR images. By jointly optimizing the similarity learning error and the segmentation error, the feature learning processes of both modalities are mutually guided. In doing so, the segmentation sub-networks are able to take advantage of the other modality's information. Considering that each modality possesses certain distinctive characteristics, we combine the higher-layer features extracted by a single-modality CNN (S-CNN) and M-CNN to form a combined CNN (C-CNN) for each modality, which is able to further utilize the complementary information of different modalities and improve the segmentation performance. The proposed M-CNN and C-CNN were evaluated on 90 CT-MR images of NPC patients. Experimental results demonstrate that our methods achieve improved segmentation performance compared to their counterparts without multi-modal information fusion and the existing CNN-based multi-modality segmentation methods.

Un-doped, metal-doped, and carbon-doped MgB2 samples were prepared by furnace cooling and quenching to investigate the second phase behavior and the resultant critical current density J c performance under different heat treatment processes, which is infrequently mentioned, and to explore the strain/lattice parameter dependence of the superconducting transition temperature. To release the residual stress, quenching induced second-phase segregation in these MgB2 samples shows a negative effect on the J c. Nevertheless, the dislocations and the lattice distortion assisted the enhancement of the high-field J c in the un-doped and metal-doped MgB2 samples, which indicated that quenching could be technically applied for the fabrication of metal-sheathed MgB2 wires and tapes to obtain excellent J c. After evaluating the micro strain and the lattice parameters' (c and a for hexagonal lattice) variation, a dome was observed in the illustration of the strain/lattice parameter c/a dependence of T c, which differed from the reported linear relation in previous work. This suggests that the c/a ratio and the strain may be the predominant parameters for scaling of the superconducting dome width in the superconducting phase diagram of MgB2.

We report the fabrication of a magnesium diboride (MgB2) coil and evaluate its persistent-mode operation in a system cooled by a cryocooler with solid nitrogen (SN2) as a cooling medium. The main purpose of SN2 was to increase enthalpy of the cold mass. For this work, an in situ processed carbon-doped MgB2 wire was used. The coil was wound on a stainless steel former in a single layer (22 turns), with an inner diameter of 109 mm and height of 20 mm without any insulation. The two ends of the coil were then joined to make a persistent-current switch to obtain the persistent-current mode. After a heat treatment, the whole coil was installed in the SN2 chamber. During operation, the resultant total circuit resistance was estimated to be

High-performance superconducting joints are essential for realizing persistent-mode magnets. Herein, we propose a concept and fabrication of such superconducting joints, which yielded reliable performance in the operating temperature range of 4.2–25 K. MgB2–MgB2 joints in magnets are known to result in deterioration of localized electrical, thermal, and mechanical properties. To overcome these problems, the ends of the two wires are inserted into a pellet press, which is then filled with a mixture of unreacted magnesium and boron powders, followed by heat treatment. The critical current capacity and joint resistance were precisely evaluated by the standard four-probe method in open-circuit and by field-decay measurements in a closed-loop, respectively. These joints demonstrated up to 66% of the current-carrying capacity of unjoined wire at 20 K, 2 T and joint resistance of 1.4 × 10−12 Ω at 4.2 K in self-field.

High performance MgB2 bulks using carbon-coated amorphous boron as a boron precursor were fabricated by Cu-activated sintering at low temperature (600 °C, below the Mg melting point). Dense nano-MgB2 grains with a high level of homogeneous carbon doping were formed in these MgB2 samples. This type of microstructure can provide a stronger flux pinning force, together with depressed volatility and oxidation of Mg owing to the low-temperature Cu-activated sintering, leading to a significant improvement of critical current density (Jc) in the as-prepared samples. In particular, the value of Jc for the carbon-coated (Mg1.1B2)Cu0.05 sample prepared here is even above 1 × 105 A cm−2 at 20 K, 2 T. The results herein suggest that the combination of low-temperature Cu-activated sintering and employment of carbon-coated amorphous boron as a precursor could be a promising technique for the industrial production of practical MgB2 bulks or wires with excellent Jc, as the carbon-coated amorphous boron powder can be produced commercially at low cost, while the addition of Cu is very convenient and inexpensive.

In the present study, MgB2 bulk was successfully synthesized by Sn-activated sintering at 600 °C for just 5 h, exhibiting a dramatic decrease in the sintering time compared to that required for sintering an undoped sample at the same temperature. According to the thermal analysis, a local Mg–Sn liquid formed first during the sintering process and then this prompted the diffusion of Mg into B. As a result, the solid–solid reaction between Mg and B was accelerated and the phase formation of MgB2 was completely achieved at low temperature. The complete formation of the MgB2 phase results in a high Jc of the Sn-doped samples sintered at low temperature. Besides this, the small MgB2 grains and nanoimpurities in the sintered samples also contributed to the good performance as regards Jc.

The sintering process of MgB2 superconductors with milled amorphous B as precursor powder was investigated using thermal analysis, and the corresponding superconductive properties were measured. It was found that the presence of a B2O3 layer on the surface of milled B particles could effectively hinder the solid–solid reaction between Mg and B during the sintering process. Since the formation of MgB2 phase mainly completes after Mg melting, the size and homogeneity of MgB2 grains mainly depends on that of the B particles. As a result, the milling treatment of B powder could result in the formation of refined and homogeneous MgB2 grains, which are responsible for the excellent Jc performance in the prepared sample with briefly milled B powder. However, as the milling time was increased, the B2O3 impurity in the B original powder increased and resulted in an obvious increase of MgO impurity in the prepared sample. These excess MgO impurities led to a significant decrease in Jc at low fields in the sample prepared with the heavily milled B powder. On the other hand, they could also serve as effective flux pinning centers due to their nanometer size and lead to an improvement in the Jc of the samples prepared with the heavily milled B powder.