A5046543349- AI in cancer detection
- Medical Image Segmentation Techniques
- Radiomics and Machine Learning in Medical Imaging
- Ultrasound Imaging and Elastography
- Advanced Neural Network Applications
- Domain Adaptation and Few-Shot Learning
- Brain Tumor Detection and Classification
- Functional Brain Connectivity Studies
- Dementia and Cognitive Impairment Research
- Digital Imaging for Blood Diseases
- Image and Signal Denoising Methods
- Photoacoustic and Ultrasonic Imaging
- Nanoplatforms for cancer theranostics
- COVID-19 diagnosis using AI
- Retinal Imaging and Analysis
- Ultrasound and Hyperthermia Applications
- Fetal and Pediatric Neurological Disorders
- Cutaneous Melanoma Detection and Management
- Image Retrieval and Classification Techniques
- Bioinformatics and Genomic Networks
- Advanced MRI Techniques and Applications
- Machine Learning in Healthcare
- Colorectal Cancer Screening and Detection
- Ultrasonics and Acoustic Wave Propagation
- Neonatal and fetal brain pathology
Shenzhen University Health Science Center
2016-2025
Shenzhen University
2016-2025
Peking University Third Hospital
2025
Peking University
2021-2025
Sun Yat-sen University
2024
China University of Petroleum, Beijing
2024
ETH Zurich
2023-2024
University of Science and Technology of China
2024
Shenyang Institute of Automation
2024
Chinese Academy of Sciences
2024
Abstract Glucose is a key energy supplier and nutrient for tumor growth. Herein, inspired by the glucose oxidase (GOx)‐assisted conversion of into gluconic acid toxic H 2 O , novel treatment paradigm starving‐like therapy developed significant tumor‐killing effects, more effective than conventional starving only cutting off supply. Furthermore, generated acidic can oxidize l ‐Arginine ( ‐Arg) NO enhanced gas therapy. By using hollow mesoporous organosilica nanoparticle (HMON) as...
This paper presents a new supervised method for vessel segmentation in retinal images. remolds the task of as problem cross-modality data transformation from image to map. A wide and deep neural network with strong induction ability is proposed model transformation, an efficient training strategy presented. Instead single label center pixel, can output map all pixels given patch. Our approach outperforms reported state-of-the-art methods terms sensitivity, specificity accuracy. The result...
Automatic localization of the standard plane containing complicated anatomical structures in ultrasound (US) videos remains a challenging problem. In this paper, we present learning-based approach to locate fetal abdominal (FASP) US by constructing domain transferred deep convolutional neural network (CNN). Compared with previous works based on low-level features, our is able represent appearance FASP and hence achieve better classification performance. More importantly, order reduce...
In this paper, a multiscale convolutional network (MSCN) and graph-partitioning-based method is proposed for accurate segmentation of cervical cytoplasm nuclei. Specifically, deep learning via the MSCN explored to extract scale invariant features, then, segment regions centered at each pixel. The coarse refined by an automated graph partitioning based on pretrained feature. texture, shape, contextual information target objects are learned localize appearance distinctive boundary, which also...
Cancer cells resist to the host immune antitumor response via multiple suppressive mechanisms, including overexpression of PD-L1 that exhausts antigen-specific CD8+ T through PD-1 receptors. Checkpoint blockade antibodies against or have shown unprecedented clinical responses. However, limited rate underlines need develop alternative engineering approaches. Here, engineered cellular nanovesicles (NVs) presenting receptors on their membranes, which enhance responses by disrupting PD-1/PD-L1...
In this paper, we present a novel framework for dermoscopy image recognition via both deep learning method and local descriptor encoding strategy. Specifically, representations of rescaled are first extracted very residual neural network pretrained on large natural dataset. Then these descriptors aggregated by orderless visual statistic features based Fisher vector (FV) to build global representation. Finally, the FV encoded used classify melanoma images using support machine with...
Accurate segmentation of cervical cells in Pap smear images is an important step automatic pre-cancer identification the uterine cervix. One major challenges overlapping cytoplasm, which has not been well-addressed previous studies. To tackle issue, this paper proposes a learning-based method with robust shape priors to segment individual cell support monitoring changes cells, vital prerequisite early detection cancer. We define splitting problem as discrete labeling task for multiple...
The quality of ultrasound (US) images for the obstetric examination is crucial accurate biometric measurement. However, manual control a labor intensive process and often impractical in clinical setting. To improve efficiency alleviate measurement error caused by improper US scanning operation slice selection, computerized fetal image assessment (FUIQA) scheme proposed to assist implementation examination. FUIQA realized with two deep convolutional neural network models, which are denoted as...
Integration of magnetic resonance imaging (MRI) and other modalities is promising to furnish complementary information for accurate cancer diagnosis imaging-guided therapy. However, most gadolinium (Gd)-chelator MR contrast agents are limited by their relatively low relaxivity high risk released-Gd-ions-associated toxicity. Herein, a radionuclide-64 Cu-labeled doxorubicin-loaded polydopamine (PDA)-gadolinium-metallofullerene core-satellite nanotheranostic agent (denoted as CDPGM) developed...
Alzheimer's disease (AD) is an irreversible progressive neurodegenerative disorder. Mild cognitive impairment (MCI) the prodromal state of AD, which further classified into a (i.e., pMCI) and stable sMCI). With development deep learning, convolutional neural networks (CNNs) have made great progress in image recognition using magnetic resonance imaging (MRI) positron emission tomography (PET) for AD diagnosis. However, due to limited availability these data, it still challenging effectively...
Automatic prostate segmentation in transrectal ultrasound (TRUS) images is of essential importance for image-guided interventions and treatment planning. However, developing such automatic solutions remains very challenging due to the missing/ambiguous boundary inhomogeneous intensity distribution TRUS, as well large variability shapes. This paper develops a novel 3D deep neural network equipped with attention modules better TRUS by fully exploiting complementary information encoded...
ABUS, or Automated breast ultrasound, is an innovative and promising method of screening for examination. Comparing to common B-mode 2D ABUS attains operator-independent image acquisition also provides 3D views the whole breast. Nonetheless, reviewing images particularly time-intensive errors by oversight might occur. For this study, we offer convolutional network, which used automated cancer detection, in order accelerate meanwhile obtain high detection sensitivity with low false positives...
Recently, deep convolutional neural networks (CNNs) have provided us an effective tool for automated polyp segmentation in colonoscopy images. However, most CNN-based methods do not fully consider the feature interaction among different layers and often cannot provide satisfactory performance. In this article, a novel attention-guided pyramid context network (APCNet) is proposed accurate robust Specifically, considering that represent aspects, APCNet first extracts multilayer features...
<h3>Context</h3> Sexually transmitted diseases are increasing rapidly in China. Surveillance data imperfectly indicate current prevalence and risk factors. <h3>Objectives</h3> To estimate the of genital chlamydial gonococcal infections to describe patterns infection by subgroup behavioral patterns. <h3>Design, Setting, Participants</h3> A national stratified probability sample 3426 Chinese individuals (1738 women 1688 men) aged 20 64 years, who were interviewed between August 1999 2000,...
CONTEXT: Intimate partner violence has been studied in many developed and developing countries.China remains one of the few large societies for which prevalence correlates intimate are unknown. METHODS: Data from a nationally representative sample women men aged 20-64 with spouse or other steady provide estimates China. Binomial multinomial logistic regression analyses adjusted design examine risk factors negative outcomes associated violence.RESULTS: Altogether, 34% 18% had ever hit during...
Convolutional Neural Networks (CNNs) have gained remarkable success in computer vision, which is mostly owe to their ability that enables learning rich image representations from large-scale annotated data. In the field of medical analysis, large amounts data may be not always available. The number acquired ground-truth sometimes insufficient train CNNs without overfitting and convergence issues scratch. Hence application deep a challenge imaging domain. However, transfer techniques are...
Automatic delineation of skin lesion contours from dermoscopy images is a basic step in the process diagnosis and treatment lesions. However, it challenging task due to high variation appearances sizes In order deal with such challenges, we propose new dense deconvolutional network (DDN) for segmentation based on residual learning. Specifically, proposed consists layers (DDLs), chained pooling (CRP), hierarchical supervision (HS). First, unlike traditional layers, DDLs are adopted maintain...