MyoPS: A benchmark of myocardial pathology segmentation combining three-sequence cardiac magnetic resonance images.

Assessment of myocardial viability is essential in diagnosis and treatment management of patients suffering from myocardial infarction, and classification of pathology on the myocardium is the key to this assessment. This work defines a new task of medical image analysis, i.e., to perform myocardial pathology segmentation (MyoPS) combining three-sequence cardiac magnetic resonance (CMR) images, which was first proposed in the MyoPS challenge, in conjunction with MICCAI 2020. Note that MyoPS refers to both myocardial pathology segmentation and the challenge in this paper. The challenge provided 45 paired and pre-aligned CMR images, allowing algorithms to combine the complementary information from the three CMR sequences for pathology segmentation. In this article, we provide details of the challenge, survey the works from fifteen participants and interpret their methods according to five aspects, i.e., preprocessing, data augmentation, learning strategy, model architecture and post-processing. In addition, we analyze the results with respect to different factors, in order to examine the key obstacles and explore the potential of solutions, as well as to provide a benchmark for future research. The average Dice scores of submitted algorithms were 0.614±0.231 and 0.644±0.153 for myocardial scars and edema, respectively. We conclude that while promising results have been reported, the research is still in the early stage, and more in-depth exploration is needed before a successful application to the clinics. MyoPS data and evaluation tool continue to be publicly available upon registration via its homepage (www.sdspeople.fudan.edu.cn/zhuangxiahai/0/myops20/).

Low-Dose CT Image Synthesis for Domain Adaptation Imaging Using a Generative Adversarial Network With Noise Encoding Transfer Learning.

Deep learning (DL) based image processing methods have been successfully applied to low-dose x-ray images based on the assumption that the feature distribution of the training data is consistent with that of the test data. However, low-dose computed tomography (LDCT) images from different commercial scanners may contain different amounts and types of image noise, violating this assumption. Moreover, in the application of DL based image processing methods to LDCT, the feature distributions of LDCT images from simulation and clinical CT examination can be quite different. Therefore, the network models trained with simulated image data or LDCT images from one specific scanner may not work well for another CT scanner and image processing task. To solve such domain adaptation problem, in this study, a novel generative adversarial network (GAN) with noise encoding transfer learning (NETL), or GAN-NETL, is proposed to generate a paired dataset with a different noise style. Specifically, we proposed a method to perform noise encoding operator and incorporate it into the generator to extract a noise style. Meanwhile, with a transfer learning (TL) approach, the image noise encoding operator transformed the noise type of the source domain to that of the target domain for realistic noise generation. One public and two private datasets are used to evaluate the proposed method. Experiment results demonstrated the feasibility and effectiveness of our proposed GAN-NETL model in LDCT image synthesis. In addition, we conduct additional image denoising study using the synthesized clinical LDCT data, which verified the merit of the proposed synthesis in improving the performance of the DL based LDCT processing method.

Incorporation of residual attention modules into two neural networks for low-dose CT denoising.

PURPOSE: The low-dose computed tomography (CT) imaging can reduce the damage caused by x-ray radiation to the human body. However, low-dose CT images have a different degree of artifacts than conventional CT images, and their resolution is lower than that of conventional CT images, which can affect disease diagnosis by clinicians. Therefore, methods for noise-level reduction and resolution improvement in low-dose CT images have inevitably become a research hotspot in the field of low-dose CT imaging. METHODS: In this paper, residual attention modules (RAMs) are incorporated into the residual encoder-decoder convolutional neural network (RED-CNN) and generative adversarial network with Wasserstein distance (WGAN) to learn features that are beneficial to improving the performances of denoising networks, and developed models are denoted as RED-CNN-RAM and WGAN-RAM, respectively. In detail, RAM is composed of a multi-scale convolution module and an attention module built on the residual network architecture, where the attention module consists of a channel attention module and a spatial attention module. The residual network architecture solves the problem of network degradation with increased network depth. The function of the attention module is to learn which features are beneficial to reduce the noise level of low-dose CT images to reduce the loss of detail in the final denoising images, which is also the key point of the proposed algorithms. RESULTS: To develop a robust network for low-dose CT image denoising, multidose-level torso phantom images provided by a cooperating equipment vendor are used to train the network, which can improve the network's adaptability to clinical application. In addition, a clinical dataset is used to test the network's migration capabilities and clinical applicability. The experimental results demonstrate that these proposed networks can effectively remove noise and artifacts from multidose CT scans. Subjective and objective analyses of multiple groups of comparison experiments show that the proposed networks achieve good noise suppression performance while preserving the image texture details. CONCLUSION: In this study, two deep learning network models are developed using multidose-level CT images acquired from a commercial spiral CT scanner. The two network models can reduce and even remove streaking artifacts, and noise from low-dose CT images confirms the effectiveness of the proposed algorithms.

Improved low-dose positron emission tomography image reconstruction using deep learned prior.

Positron emission tomography (PET) is a promising medical imaging technology that provides non-invasive and quantitative measurement of biochemical process in the human bodies. PET image reconstruction is challenging due to the ill-poseness of the inverse problem. With lower statistics caused by the limited detected photons, low-dose PET imaging leads to noisy reconstructed images with much quality degradation. Recently, deep neural networks (DNN) have been widely used in computer vision tasks and attracted growing interests in medical imaging. In this paper, we proposed a maximuma posteriori(MAP) reconstruction algorithm incorporating a convolutional neural network (CNN) representation in the formation of the prior. Rather than using the CNN in post-processing, we embedded the neural network in the reconstruction framework for image representation. Using the simulated data, we first quantitatively evaluated our proposed method in terms of the noise-bias tradeoff, and compared with the filtered maximum likelihood (ML), the conventional MAP, and the CNN post-processing methods. In addition to the simulation experiments, the proposed method was further quantitatively validated on the acquired patient brain and body data with the tradeoff between noise and contrast. The results demonstrated that the proposed CNN-MAP method improved noise-bias tradeoff compared with the filtered ML, the conventional MAP, and the CNN post-processing methods in the simulation study. For the patient study, the CNN-MAP method achieved better noise-contrast tradeoff over the other three methods. The quantitative enhancements indicate the potential value of the proposed CNN-MAP method in low-dose PET imaging.

Strategies to approach high performance in Cr3+-doped phosphors for high-power NIR-LED light sources.

Broadband near-infrared (NIR)-emitting phosphors are key for next-generation smart NIR light sources based on blue LEDs. To achieve excellent NIR phosphors, we propose a strategy of enhancing the crystallinity, modifying the micromorphology, and maintaining the valence state of Cr3+ in Ca3Sc2Si3O12 garnet (CSSG). By adding fluxes and sintering in a reducing atmosphere, the internal quantum efficiency (IQE) is greatly enhanced to 92.3%. The optimized CSSG:6%Cr3+ exhibits excellent thermal stability. At 150 °C, 97.4% of the NIR emission at room temperature can be maintained. The fabricated NIR-LED device emits a high optical power of 109.9 mW at 520 mA. The performances of both the achieved phosphor and the NIR-LED are almost the best results until now. The mechanism for the optimization is investigated. An application of the NIR-LED light source is demonstrated.

Electron-phonon coupling mechanisms of broadband near-infrared emissions from Cr3+ in the Ca3Sc2Si3O12 garnet.

Cr3+ in the Ca3Sc2Si3O12 garnet (CSSG) has the ability to convert blue light to broadband near-infrared (NIR) emissions, which is a promising strategy for next-generation smart NIR light sources based on blue LEDs. The Cr3+ luminescence strongly depends on temperature due to electron-phonon coupling (EPC). We reveal the EPC mechanism of Cr3+ in CSSG for the first time by temperature-dependent photoluminescence measurement from 77 to 573 K and cathodoluminescence using a scanning electron microscope. Cr3+ occupies the Sc3+ site and experiences a weak crystal field in CSSG, manifesting a broad NIR emission in the 700-900 nm range that originates from the 4T2g→4A2g transition. The zero phonon line (ZPL) of the 4T2 state is observed at ∼713 nm with a vibrational energy of ∼310 cm-1. A strong EPC leads to a large Stokes shift (∼2900 cm-1). The Huang-Rhys parameter (S = 4), crystal field strength (Dq/B), and Racah parameters (B and C) are estimated.

YAG:Ce3+ Transparent Ceramic Phosphors Brighten the Next-Generation Laser-Driven Lighting.

Y3 Al5 O12 :Ce3+ (YAG:Ce3+ ) transparent ceramic phosphors (TCPs) are regarded as the most promising luminescent converter for laser-driven (LD) lighting. High-quality YAG:Ce3+ TCPs are still urgent for high efficiency LD lighting devices. YAG:Ce3+ TCPs in a vacuum ambience by using nano-sized raw materials are prepared. Controlling defects by adding nano-sized MgO and SiO2 simultaneously enables a high transmittance nearly 80%. After annealing in air furthermore, the luminous efficiency is enhanced greatly from 106 to 223 lm W-1 , which is the best result reported now for LD lighting. These results demonstrate that the optimizing YAG:Ce3+ TCPs in a fitting strategy will brighten once again in the next-generation LD lighting. Based on scanning electron microscopy (SEM) coupled with a cathodoluminescence system, defects and Ce3+ distributions in grains are identified directly for the first time.

Full spectrum core-shell phosphors under ultraviolet excitation.

A YAG: Ce/MgY4Si3O13:Ce-Y2O3:Eu core-shell structure was designed and accomplished via a urea homogeneous precipitation method. The as prepared phosphors can emit photons with a broad range of wavelengths from 340 nm to 700 nm under excitation light of 330 nm. The internal quantum efficiency can reach up to 68%.

Transparent Ceramics Enabling High Luminous Flux and Efficacy for the Next-Generation High-Power LED Light.

By designing novel chemical compositions and controlling precursor powders, perfect transparent ceramics (TCs) of Gd3Al4GaO12:2%Ce3+ garnets (GAGG:2%Ce3+) were achieved for the first time. Ce3+ distributions in ceramics were revealed by micromorphology and micro-cathodoluminescence. Benefiting from the components that are rich in red emission, warm light with a correlated color temperature of about 2800 K was generated when TC was used in high-power (hp) blue light emitting diodes (LEDs) (∼450 nm, ∼34 W driven at 1 A). The hp LED device shows high brightness with a luminous flux nearly 2100 lm. The luminous efficacy even reaches 388 lm/W, which is the highest value reported till now, indicating that the use of GAGG:2%Ce3+ TCs promises energy saving. We believe that this work will open a new perspective to develop TCs for next-generation hp LED lighting to substitute traditional hp xenon lamps and high-pressure sodium lamps.

Warm White Light with a High Color-Rendering Index from a Single Gd3Al4GaO12:Ce3+ Transparent Ceramic for High-Power LEDs and LDs.

Transparent ceramics (TCs) are promising for high-power (hp) white light-emitting diode (WLED) and laser diode (LD) lighting. However, comfortable warm white light has not been achieved only using a single TC in hp-WLEDs/LDs. Herein, highly transparent Gd3Al4GaO12:Ce3+ (GAGG:Ce3+) TCs (transmittance, T = 55.9-80.2%) were prepared via a solid-state reaction. Ce3+ as a doped activator center in grains plays a positive role in luminescence based on the microstructural investigations by scanning electron microscopy and the cathodoluminescence system. T decreases upon increasing the Ce3+ concentration and/or the ceramic thickness, whereas the luminous efficacy of hp-WLEDs/LDs goes up. For blue hp-LEDs driven at 350 mA or LDs of 2 W, warm white light with a low correlated-color temperature of ∼3000 K was achieved by a single GAGG:Ce3+ TC, benefiting from its broad emission band (full width at half maximum, FWHM = 133-137 nm) and abundant red components (peaking at about 568-574 nm). The color-rendering index of hp-WLEDs reaches 78.9. These results are much better than the performance of the traditional Y3Al5O12:Ce3+ (YAG:Ce3+) TC, indicating that GAGG:Ce3+ TCs are promising color converters for hp-WLEDs/LDs with a comfortable warm white light.

A two-step solid-state reaction to synthesize the yellow persistent Gd3Al2Ga3O12:Ce3+ phosphor with an enhanced optical performance for AC-LEDs.

By using a two-step solid-state reaction to synthesize the yellow-persistent Gd3Al2Ga3O12:Ce3+ phosphor, both the luminescence and afterglow properties were greatly enhanced. Its internal quantum efficiency reaches as high as 81.9%, indicating its promising application in reducing the flicker effect in AC-LED-based white light systems.

Highly transparent cerium doped gadolinium gallium aluminum garnet ceramic prepared with precursors fabricated by ultrasonic enhanced chemical co-precipitation.

Cerium doped gadolinium gallium aluminum garnet (GGAG:Ce) ceramic precursors have been synthesized with an ultrasonic chemical co-precipitation method (UCC) and for comparison with a traditional chemical co-precipitation method (TCC). The effect of ultra-sonication on the morphology of powders and the transmittance of GGAG:Ce ceramics are studied. The results indicate that the UCC method can effectively improve the homogenization and sinterability of GGAG:Ce powders, which contribute to obtain high transparent GGAG ceramic with the highest transmittance of 81%.

A novel banded structure ceramic phosphor for high-power white LEDs.

A novel banded structure ceramic phosphor has been fabricated in this research. This structure provides a convenient and effective method for regulating the full spectrum. It has the advantage of being able to adjust the intensity of all three primary colors independently, regardless of the mutual absorption among different active ions.

Origin and Luminescence of Anomalous Red-Emitting Center in Rhombohedral Ba9Lu2Si6O24:Eu(2+) Blue Phosphor.

We obtain a blue phosphor, Ba9Lu2Si6O24:Eu(2+) (BLS:Eu(2+)), which shows a strong emission peak at 460 nm and a weak tail from 460 to 750 nm. A 610 nm red emission is observed for the first time in this kind of rhombohedral structure material, which is much different from the same crystal structure of Ba9Sc2Si6O24:Eu(2+) and Ba9Y2Si6O24:Eu(2+). The luminescence properties and decays from 10 to 550 K are discussed. The new red emission arises from a trapped exciton state of Eu(2+) at the Ba site with a larger coordination number (12-fold). It exhibits abnormal luminescence properties with a broad bandwidth and a large Stokes shift. Under the 400 nm excitation, the external quantum efficiency of BLS:Eu(2+) is 45.4%, which is higher than the 35.7% for the commercial blue phosphor BAM:Eu(2+). If the thermal stability of BLS:Eu(2+) can be improved, it will show promising applications in efficient near-UV-based white LEDs.

Near and Mid-Infrared Emission Characteristics of Er³âº/Tm³âº/Ho³âº-Doped LiYF4 Single Crystals Excited by Laser Diode.

Yttrium lithium fluoride (LiYF4) single crystals triply doped with Er³âº/Tm³âº/Ho³âº are synthesized by a vertical Bridgman method. Absorption spectra, emission spectra, and decay curves are measured to investigate the luminescent properties of the crystals. Compared with Er³âº singly doped and Er³âº/Tm³âº and Er³âº/Ho³âº doubly doped LiYF4 crystals, an intense emission around 2.7 µm can be obtained in the triply doped LiYF4 crystal under excitation of 980 nm laser diode. Meanwhile, the near infrared emission at 1.5 µm from Er³âº in the triply doped crystal is effectively reduced. The possible energy transfer processes and the luminescent mechanisms for enhancing emission at 2.7 µm and quenching emission at 1.5 µm in the Er³âº/Ho³âº/Tm³âº triply doped crystals are proposed. The large energy transfer efficiency of 82.0% and excellent optical transmission indicate that this Er³âº/Tm³âº/Ho³âº triply doped crystal can be considered as a promising material for a mid- infrared laser at 2.7 µm.

Growth and Optical Spectra of Tb³âº/Sm³âº/Ce³âº Triply Doped LiYF4Single Crystals for White-Light LEDs.

The Tb³âº/Sm³âº/Ce³âº triply doped LiYF4 single crystals were grown by a modified Bridgman method. The absorption spectra, excitation spectra, and fluorescence spectra of Tb³âº/Sm³âº/Ce³âº ions in LiYF4 crystals were measured. The fluorescence spectra of several bands, mainly located at purplish blue ~413 nm (5D3 --> 7F5), yellowish green ~542 nm (5D4 --> 7F5), and red ~643 nm (4G5/2 --> 6H9/2), were observed under excitation of ultraviolet light. White light could be generated by the mixture of the multicolor lights. The luminous intensities varied slightly with the excitation wavelength from 300 nm to 400 nm and doping Tb³âº/Sm³âº/Ce³âº ion concentration. The chromatic- ity coordinates of the crystal could be modified by changing the excitation wavelengths and the concentrations of Tb³âº/Sm³âº/Ce³âº ions. A near-ideal white light emission could be obtained from 1.25 mol% Tb³âº, 1.98 mol% Sm³âº, 0.44 mol% Ce³âº triply doped LiYF4 single crystal with chro- maticity coordinates of x = 0.3090, y = 0.3223, color temperature Tc = 6777 K, color rendering index Ra = 77 and color quality scale Qa = 76 under the excitation of a 374 nm light.

Energy Transfer and 1.8 µm Emission in Tm³âº/Yb³âº Co-Doped LiYF4Crystal.

LiYF4 single crystals co-doped with various Tm³âº/Yb³âº concentrations were grown using the Bridg- man method. The luminescent properties of the crystals were investigated through emission spectra, emission cross section, and decay curves under excitation by 980 nm. Compared with the Tm³âº single-doped LiYF4 crystal, an enhanced emission band from 1600 to 2150 nm was observed upon excitation of a 980 nm laser diode. The energy transfer from Yb³âº to Tm³âº and the optimum fluo- rescence emission around 1.80 µm of Tm³âº ion were investigated. The maximum emission cross section at 1.8 µm was calculated to be 1.48 x 10⁻²° cm² according to the measured absorption spectrum. The high energy transfer efficiency of 86.5% from Yb³âº to Tm³âº ion demonstrate that the Yb³âº ions can efficiently sensitize the Tm³âº ions.

Effect of Yb(3+) on the Crystal Structural Modification and Photoluminescence Properties of GGAG:Ce(3+).

Gadolinium gallium aluminum garnet (GGAG) is a very promising host for the highly efficient luminescence of Ce(3+) and shows potential in radiation detection applications. However, the thermodynamically metastable structure would be slanted against it from getting high transparency. To stabilize the crystal structure of GGAG, Yb(3+) ions were codoped at the Gd(3+) site. It is found that the decomposition of garnet was suppressed and the transparency of GGAG ceramic was evidently improved. Moreover, the photoluminescence of GGAG:Ce(3+),xYb(3+) with different Yb(3+) contents has been investigated. When the Ce(3+) ions were excited under 475 nm, a typical near-infrared region emission of Yb(3+) ions can be observed, where silicon solar cells have the strongest absorption. Basing on the lifetimes of Ce(3+) ions in the GGAG:Ce(3+),xYb(3+) sample, the transfer efficiency from Ce(3+) to Yb(3+) and the theoretical internal quantum efficiency can be calculated and reach up to 86% and 186%, respectively. This would make GGAG:Ce(3+),Yb(3+) a potential attractive downconversion candidate for improving the energy conversion efficiency of crystalline silicon (c-Si) solar cells.

A first-principles study on the phonon transport in layered BiCuOSe.

First-principles calculations are employed to investigate the phonon transport of BiCuOSe. Our calculations reproduce the lattice thermal conductivity of BiCuOSe. The calculated grüneisen parameter is 2.4 ~ 2.6 at room temperature, a fairly large value indicating a strong anharmonicity in BiCuOSe, which leads to its ultralow lattice thermal conductivity. The contribution to total thermal conductivity from high-frequency optical phonons, which are mostly contributed by the vibrations of O atoms, is larger than 1/3, remarkably different from the usual picture with very little contribution from high-frequency optical phonons. Our calculations show that both the high group velocities and low scattering processes involved make the high-frequency optical modes contribute considerably to the total lattice thermal conductivity. In addition, we show that the sound velocity and bulk modulus along a and c axes exhibit strong anisotropy, which results in the anisotropic thermal conductivity in BiCuOSe.

High thermoelectric performance in two-dimensional graphyne sheets predicted by first-principles calculations.

The thermoelectric properties of two-dimensional graphyne sheets are investigated by using first-principles calculations and the Boltzmann transport equation method. The electronic structure indicates a semiconducting phase for graphyne, compared with the metallic phase of graphene. Consequently, the obtained Seebeck coefficient and the power factor of graphyne are much higher than those of graphene. The calculated phonon mean free path for graphene is 866 nm, which is in good agreement with the experimental value of 775 nm. Meanwhile the phonon mean free path of graphyne is only 60 nm, leading to two order lower thermal conductivity than graphene. We show that the low thermal conductivity of graphyne is due to its mixed sp/sp(2) bonding. Our calculations show that the optimized ZT values of graphyne sheets can reach 5.3 at intermediate temperature by appropriate doping.