This paper presents a new supervised method for vessel segmentation in retinal images. This method remolds the task of segmentation as a problem of cross-modality data transformation from retinal ...image to vessel map. A wide and deep neural network with strong induction ability is proposed to model the transformation, and an efficient training strategy is presented. Instead of a single label of the center pixel, the network can output the label map of all pixels for a given image patch. Our approach outperforms reported state-of-the-art methods in terms of sensitivity, specificity and accuracy. The result of cross-training evaluation indicates its robustness to the training set. The approach needs no artificially designed feature and no preprocessing step, reducing the impact of subjective factors. The proposed method has the potential for application in image diagnosis of ophthalmologic diseases, and it may provide a new, general, high-performance computing framework for image segmentation.
•High glucose causes blood–brain barrier disruption in the early reperfusion of stroke.•MMP-2/9-mediated extracellular ZO-1 degradation.•Caveolin-1-mediated intracellular translocation of ...ZO-1.•Autophagy-lysosome-mediated ZO-1 degradation.
Post-stroke hyperglycemia during early reperfusion increases blood–brain barrier (BBB) permeability and subsequently aggravates brain injury and clinical prognosis. The decreased level of tight junction proteins (TJPs) has been reported but the underlying mechanism remains largely elusive. Herein we designed to investigate the detailed molecular events in brain microvascular endothelial cells (BMECs) ex and in vivo. After oxygen–glucose deprivation (OGD) for 90 min and reperfusion with 8 or 16 mM glucose for 30 min, glucose at 16 mM caused significant decrease in the TJP expression and particularly ZO-1 redistribution from membrane to cytoplasm of BMECs. High glucose also markedly promoted the secretion of MMP-2/9 and oxidative/nitrosative stress, enhanced autophagy and increased the Caveolin-1 and LAMP-2 expression. Moreover, in vivo experiments demonstrated that rapamycin-enhanced autophagy further caused ZO-1 reduction and the increased BBB permeability. Therefore, high-glucose exposure in the early reperfusion causes the BBB disruption, with MMP-2/9-mediated extracellular degradation, caveolin-1-mediated intracellular translocation and autophagy-lysosome-mediated degradation of ZO-1 protein all together involved in the process. The role of MMP-2/-9 and autophagy in the modulation of paracellular permeability was confirmed by pharmacological inhibition. Therefore, our findings may provide new insights into targeting ZO-1 regulation for the purpose of significantly improving the clinical prognosis of ischemic stroke.
In this paper, a multiscale convolutional network (MSCN) and graph-partitioning-based method is proposed for accurate segmentation of cervical cytoplasm and nuclei. Specifically, deep learning via ...the MSCN is explored to extract scale invariant features, and then, segment regions centered at each pixel. The coarse segmentation is refined by an automated graph partitioning method based on the pretrained feature. The texture, shape, and contextual information of the target objects are learned to localize the appearance of distinctive boundary, which is also explored to generate markers to split the touching nuclei. For further refinement of the segmentation, a coarse-to-fine nucleus segmentation framework is developed. The computational complexity of the segmentation is reduced by using superpixel instead of raw pixels. Extensive experimental results demonstrate that the proposed cervical nucleus cell segmentation delivers promising results and outperforms existing methods.
Photoacoustic imaging (PAI), a novel imaging modality based on photoacoustic effect, shows great promise in biomedical applications. By converting pulsed laser excitation into ultrasonic emission, ...PAI combines the advantages of optical imaging and ultrasound imaging, which benefits rich contrast, high resolution and deep tissue penetration. In this paper, we introduced recent advances of contrast agents, applications, and signal enhancement strategies for PAI. The PA contrast agents were categorized by their components, mainly including inorganic and organic PA contrast agents. The applications of PAI were summarized as follows: deep tumor imaging, therapeutic responses monitoring, metabolic imaging, pH detection, enzyme detection, reactive oxygen species (ROS) detection, metal ions detection, and so on. The enhancement strategies of PA signals were highlighted. In the end, we elaborated on the challenges and provided perspectives of PAI for deep-tissue biomedical applications.
Glucose is a key energy supplier and nutrient for tumor growth. Herein, inspired by the glucose oxidase (GOx)‐assisted conversion of glucose into gluconic acid and toxic H2O2, a novel treatment ...paradigm of starving‐like therapy is developed for significant tumor‐killing effects, more effective than conventional starving therapy by only cutting off the energy supply. Furthermore, the generated acidic H2O2 can oxidize l‐Arginine (l‐Arg) into NO for enhanced gas therapy. By using hollow mesoporous organosilica nanoparticle (HMON) as a biocompatible/biodegradable nanocarrier for the co‐delivery of GOx and l‐Arg, a novel glucose‐responsive nanomedicine (l‐Arg‐HMON‐GOx) has been for the first time constructed for synergistic cancer starving‐like/gas therapy without the need of external excitation, which yields a remarkable H2O2–NO cooperative anticancer effect with minimal adverse effect.
Synergistic therapy: A biocompatible and biodegradable nanomedicine based on glucose oxidase/l‐arginine co‐loaded hollow mesoporous organosilica nanoparticles has been successfully constructed for converting intratumoral glucose into high concentrations of toxic hydrogen peroxide and nitric oxide. An endogenous synergistic cancer therapy for efficient tumor eradication has been developed.
Cancer cells resist to the host immune antitumor response via multiple suppressive mechanisms, including the overexpression of PD‐L1 that exhausts antigen‐specific CD8+ T cells through PD‐1 ...receptors. Checkpoint blockade antibodies against PD‐1 or PD‐L1 have shown unprecedented clinical responses. However, limited host response rate underlines the need to develop alternative engineering approaches. Here, engineered cellular nanovesicles (NVs) presenting PD‐1 receptors on their membranes, which enhance antitumor responses by disrupting the PD‐1/PD‐L1 immune inhibitory axis, are reported. PD‐1 NVs exhibit a long circulation and can bind to the PD‐L1 on melanoma cancer cells. Furthermore, 1‐methyl‐tryptophan, an inhibitor of indoleamine 2,3‐dioxygenase can be loaded into the PD‐1 NVs to synergistically disrupt another immune tolerance pathway in the tumor microenvironment. Additionally, PD‐1 NVs remarkably increase the density of CD8+ tumor infiltrating lymphocytes in the tumor margin, which directly drive tumor regression.
Cellular nanovesicles (NVs) presenting PD‐1 receptors on their membrane are genetically engineered for disturbing the PD1/PD‐L1 immune inhibitory axis. Additionally, 1‐methyl‐tryptophan (1‐MT), an inhibitor of indoleamine 2,3‐dioxygenase (IDO) can be loaded into the PD‐1 NVs to synergistically promote antitumor efficacy. This formulation provides a promising strategy that leverages functions of immune checkpoint blockade and encapsulated therapeutics for enhancing cancer immunotherapy.
The global utilization of H2O2 is currently around 4 million tons per year and is expected to continue to increase in the future. H2O2 is mainly produced by the anthraquinone process, which involves ...multiple steps in terms of alkylanthraquinone hydrogenation/oxidation in organic solvents and liquid–liquid extraction of H2O2. The energy‐intensive and environmentally unfriendly anthraquinone process does not meet the requirements of sustainable and low‐carbon development. The electrocatalytic two‐electron (2 e−) oxygen reduction reaction (ORR) driven by renewable energy (e.g. solar and wind power) offers a more economical, low‐carbon, and greener route to produce H2O2. However, continuous and decentralized H2O2 electrosynthesis still poses many challenges. This Minireview first summarizes the development of devices for H2O2 electrosynthesis, and then introduces each component, the assembly process, and some optimization strategies.
Electrochemical reactors for continuous decentralized H2O2 production are described in this Minireview, with separate discussions of flow field plates, catalyst layers, gas diffusion layers, membranes, shapes, and electrolyte compartments. The key factors of these parts and the optimization strategies for assembling flow cells are summarized. Insights and perspectives on key components are given.
Ultrasound (US) has become one of the most commonly performed imaging modalities in clinical practice. It is a rapidly evolving technology with certain advantages and with unique challenges that ...include low imaging quality and high variability. From the perspective of image analysis, it is essential to develop advanced automatic US image analysis methods to assist in US diagnosis and/or to make such assessment more objective and accurate. Deep learning has recently emerged as the leading machine learning tool in various research fields, and especially in general imaging analysis and computer vision. Deep learning also shows huge potential for various automatic US image analysis tasks. This review first briefly introduces several popular deep learning architectures, and then summarizes and thoroughly discusses their applications in various specific tasks in US image analysis, such as classification, detection, and segmentation. Finally, the open challenges and potential trends of the future application of deep learning in medical US image analysis are discussed.
Integration of magnetic resonance imaging (MRI) and other imaging modalities is promising to furnish complementary information for accurate cancer diagnosis and imaging‐guided therapy. However, most ...gadolinium (Gd)–chelator MR contrast agents are limited by their relatively low relaxivity and high risk of released‐Gd‐ions‐associated toxicity. Herein, a radionuclide‐64Cu‐labeled doxorubicin‐loaded polydopamine (PDA)–gadolinium‐metallofullerene core–satellite nanotheranostic agent (denoted as CDPGM) is developed for MR/photoacoustic (PA)/positron emission tomography (PET) multimodal imaging‐guided combination cancer therapy. In this system, the near‐infrared (NIR)‐absorbing PDA acts as a platform for the assembly of different moieties; Gd3N@C80, a kind of gadolinium metallofullerene with three Gd ions in one carbon cage, acts as a satellite anchoring on the surface of PDA. The as‐prepared CDPGM NPs show good biocompatibility, strong NIR absorption, high relaxivity (r
1 = 14.06 mM−1 s−1), low risk of release of Gd ions, and NIR‐triggered drug release. In vivo MR/PA/PET multimodal imaging confirms effective tumor accumulation of the CDPGM NPs. Moreover, upon NIR laser irradiation, the tumor is completely eliminated with combined chemo‐photothermal therapy. These results suggest that the CDPGM NPs hold great promise for cancer theranostics.
A versatile cancer theranostic agent, with good biocompatibility, high relaxivity, low risk of release of Gd ions, and near‐infrared‐triggered drug release, is developed based on core–satellite polydopamine–gadolinium metallofullerene, for magnetic resonance/photoacoustic/positron emission tomography multimodal imaging and chemo‐photothermal combination therapy.
Automatic localization of the standard plane containing complicated anatomical structures in ultrasound (US) videos remains a challenging problem. In this paper, we present a learning-based approach ...to locate the fetal abdominal standard plane (FASP) in US videos by constructing a domain transferred deep convolutional neural network (CNN). Compared with previous works based on low-level features, our approach is able to represent the complicated appearance of the FASP and hence achieve better classification performance. More importantly, in order to reduce the overfitting problem caused by the small amount of training samples, we propose a transfer learning strategy, which transfers the knowledge in the low layers of a base CNN trained from a large database of natural images to our task-specific CNN. Extensive experiments demonstrate that our approach outperforms the state-of-the-art method for the FASP localization as well as the CNN only trained on the limited US training samples. The proposed approach can be easily extended to other similar medical image computing problems, which often suffer from the insufficient training samples when exploiting the deep CNN to represent high-level features.