A5100637983- Medical Imaging Techniques and Applications
- Radiomics and Machine Learning in Medical Imaging
- Medical Image Segmentation Techniques
- Plasma Applications and Diagnostics
- Plasma Diagnostics and Applications
- Advanced Neural Network Applications
- Aerosol Filtration and Electrostatic Precipitation
- Advanced Image and Video Retrieval Techniques
- Electrohydrodynamics and Fluid Dynamics
- Advanced MRI Techniques and Applications
- Prostate Cancer Treatment and Research
- AI in cancer detection
- Medical Imaging and Analysis
- Robotics and Sensor-Based Localization
- Electric Motor Design and Analysis
- Cerebrovascular and Carotid Artery Diseases
- Face and Expression Recognition
- COVID-19 diagnosis using AI
- Remote Sensing and LiDAR Applications
- Human Pose and Action Recognition
- Applied Advanced Technologies
- DNA and Biological Computing
- Respiratory viral infections research
- Image and Video Quality Assessment
- Machine Learning and ELM
Southwest Petroleum University
2025
PetroChina Southwest Oil and Gas Field Company (China)
2025
Johns Hopkins University
2019-2024
Central Academy of Drama
2023-2024
Central People's Hospital of Zhanjiang
2024
State Key Laboratory of Microbial Technology
2024
Johns Hopkins Medicine
2023-2024
Huazhong University of Science and Technology
2019-2024
Qingdao Binhai University
2024
Shandong University
2017-2024
In the last decade, convolutional neural networks (ConvNets) have dominated and achieved state-of-the-art performances in a variety of medical imaging applications. However, ConvNets are still limited by lacking understanding long-range spatial relations an image. The recently proposed Vision Transformer (ViT) for image classification uses purely self-attention-based model that learns to focus on relevant parts Nevertheless, ViT emphasizes low-resolution features because consecutive...
Position emission tomography (PET) is widely used in clinics and research due to its quantitative merits high sensitivity, but suffers from low signal-to-noise ratio (SNR). Recently convolutional neural networks (CNNs) have been improve PET image quality. Though successful efficient local feature extraction, CNN cannot capture long-range dependencies well limited receptive field. Global multi-head self-attention (MSA) a popular approach information. However, the calculation of global MSA for...
Mice that overexpress the human Cu,Zn superoxide dismutase-1 mutant G93A develop a delayed and progressive motor neuron disease similar to amyotrophic lateral sclerosis (ALS). Most current studies of therapeutics in these mice date have involved administration agents long before onset symptoms, which cannot currently be accomplished ALS patients. We examined effects manganese porphyrin AEOL 10150 (manganese [III] tetrakis[N-N'-diethylimidazolium-2-yl]porphyrin) given at symptom found, three...
In this paper we present a novel network architecture, called Multi-Scale Cascade Network (MSC-Net), to identify the most visually conspicuous objects in an image. Our consists of several stages (sub-networks) for handling saliency detection across different scales. All these sub-networks form cascade structure (in coarse-to-fine manner) where same underlying convolutional feature representations are fully shared. Compared with existing CNN-based models, MSC-Net can naturally enable learning...
Influenza A viruses pose a significant threat globally each year, underscoring the need for vaccine- or antiviral-based broad-protection strategy. Here, we describe chimeric monoclonal antibody, C12H5, that offers neutralization against seasonal and pandemic H1N1 viruses, cross-protection some H5N1 viruses. Notably, C12H5 mAb broad neutralizing activity by controlling virus entry egress, protection viral challenge in vivo. Through structural analyses, show engages hemagglutinin (HA), major...
Abstract Background Radiotherapy (RT) combined with cetuximab is the standard treatment for patients inoperable head and neck cancers. Segmentation of (H&N) tumors a prerequisite radiotherapy planning but time‐consuming process. In recent years, deep convolutional neural networks (DCNN) have become de facto automated image segmentation. However, due to expensive computational cost associated enlarging field view in DCNNs, their ability model long‐range dependency still limited, this can...
Derived from camelid heavy-chain antibodies, nanobodies (Nbs) are the smallest natural antibodies and an ideal tool in biological studies because of their simple structure, high yield, low cost. Nbs possess significant potential for developing highly specific user-friendly diagnostic assays. Despite offering considerable advantages detection applications, knowledge is limited regarding exclusive use lateral flow immunoassay (LFIA) detection. Herein, we present a novel double "Y"...
In-hospital mortality following hip fractures is a significant concern, and accurate prediction of this outcome crucial for appropriate clinical management. Nonetheless, there lack effective tools in practice. By utilizing artificial intelligence (AI) machine learning techniques, study aims to develop predictive model that can assist clinicians identifying geriatric fracture patients at higher risk in-hospital mortality.
The automatic segmentation of main vessels on X-ray angiography (XRA) images is great importance in the smart coronary artery disease diagnosis system. However, existing methods have been developed to this task, but these difficulty recognizing structure XRA images. Main vessel still a challenging task due diversity and small-size region In study, we propose robust method for by using deep learning architectures with fully convolutional networks. Four models based UNet architecture are...
Background: Multiple organ dysfunction syndrome (MODS) is common after sepsis and increases mortality. Lactate (Lac) can assess the prognosis of patients. Albumin (Alb) closely associated with inflammatory response in This work evaluated predictive value Lac/Alb for Methods: Data 160 patients were retrospectively collected. Lac Alb levels measured upon admission, at 24 hours 48 later. Using 0.45 as cutoff Lac/Alb, rolled into high-level (HL) low-level (LL) groups. MODS rates mortality...
Uptake segmentation and classification on PSMA PET/CT are important for automating whole-body tumor burden determinations. We developed evaluated an automated deep learning (DL)-based framework that segments classifies uptake PET/CT. identified 193 [18F] DCFPyL scans of patients with biochemically recurrent prostate cancer from two institutions, including 137 training internally testing, 56 another institution external testing. Two radiologists segmented labelled foci as suspicious or...
Abstract Producing spatial transformations that are diffeomorphic is a key goal in deformable image registration. As transformation should have positive Jacobian determinant $$\vert J\vert $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mo>|</mml:mo> <mml:mi>J</mml:mi> </mml:mrow> </mml:math> everywhere, the number of pixels (2D) or voxels (3D) with <0$$ <mml:mo><</mml:mo> <mml:mn>0</mml:mn> has been used to test for diffeomorphism and also measure...
The pore structure of shale is a critical factor influencing the occurrence and flow gas. Characterizing studying its heterogeneity are paramount importance for deeper understanding laws governing hydrocarbon occurrence, as well enhancing efficiency exploration development. This work addresses complex characteristics multiscale coupling in systems reservoirs, focusing on ultra-deep Qiongzhusi Formation southern region. Through integrated application cross-scale observation techniques...
With increasing sinusoidal frequency or its voltage amplitude, a central filament emerges in diffuse plume downstream of an argon jet with single-electrode geometry. The co-existing and part, also referred to as diffuse-and-filamentary (DAF) plume, is formed after sudden length increment from plume. Only one negative discharge appears per cycle for the Besides discharge, there positive DAF By fast photography, behaves plump plasma bullets propagating along flow, which results part...
A plasma jet with inert working gas operates in a streamer discharge mechanism. The propagation behavior of streamers determines the distribution active species, which play key role applications jets. To make clear under influence residual positive ions, two-dimensional fluid model based on continuity, conservation, and Poisson's equations is employed to numerically investigate dynamic when it approaches cloud ions different densities scales. Results indicate that always propagates along...
Abstract Plasma plume morphology, related to streamer behavior and active‐species distribution of plasma jets, is important for fixed‐point precise treatment workpieces. For a peculiar snake‐like plume, discharge aspects are investigated with varying argon flow rate ( Q ) in this paper. Results indicate that the length distance between two neighboring peaks increase increasing . Fast photography indicates composed positive streamers propagating along reproducible serpentine trajectory, which...
Abstract Compared with the diffuse mode of helium plasma jets, a filamentary is normally produced in plume less expensive inert gas, such as argon, which undesirable because intense discharge may cause damage to fragile samples. Many efforts have been attempted realize an argon jet. In this paper, realized atmospheric pressure jet increasing biased voltage ( V b ) applied downstream plate electrode. Results show that transits from increase . Waveforms voltage, current and integrated light...
Being able to generate a remote plasma plume, the atmospheric pressure jet has become an indispensable tool for extensive application fields. A plume usually straight column morphology, which results from straight-line or stochastic snake-like propagations of streamers. The propagation streamers is unclear in mechanism. In this paper, meandering generated first time downstream argon excited by positively biased sinusoidal voltage, transits with increasing bias voltage (Vb). Results indicate...
Plasma jet is an important low-temperature plasma source in extensive application fields. To promote the production of active oxygen species, often introduced into inert working gas. However, influence content on discharge characteristics argon not clear. Aim to this status, a single-electrode geometry employed investigate concentration ( C O ) aspects. Results indicate that with increasing (≤ 0.6%), plume transits from diffuse morphology hollow structure. Electrical and optical measurements...
For the majority of learning-based segmentation methods, a large quantity high-quality training data is required. In this paper, we present novel model that could be trained semi- or un- supervised. Specifically, in unsupervised setting, parameterize Active contour without edges (ACWE) framework via convolutional neural network (ConvNet), and optimize parameters ConvNet using self-supervised method. another setting (semi-supervised), auxiliary ground truth used during training. We show...
In the past, optimization-based registration models have used spatially-varying regularization to account for deformation variations in different image regions. However, deep learning-based mostly relied on spatially-invariant regularization. Here, we introduce an end-to-end framework that uses neural networks learn a regularizer directly from data. The hyperparameter of proposed is conditioned into network, enabling easy tuning strength. method built upon Transformer-based model, but it can...