A5021757255- Photoacoustic and Ultrasonic Imaging
- Ultrasound Imaging and Elastography
- Retinal Diseases and Treatments
- AI in cancer detection
- Ultrasonics and Acoustic Wave Propagation
- Image and Signal Denoising Methods
- Advanced Image Processing Techniques
- Cancer Immunotherapy and Biomarkers
- Retinal Imaging and Analysis
- Image Processing Techniques and Applications
- Optical Imaging and Spectroscopy Techniques
- Evolutionary Algorithms and Applications
- Ferroptosis and cancer prognosis
- Reinforcement Learning in Robotics
- 3D Shape Modeling and Analysis
- Metaheuristic Optimization Algorithms Research
- Urinary and Genital Oncology Studies
- Adversarial Robustness in Machine Learning
- Neuroinflammation and Neurodegeneration Mechanisms
- Radiomics and Machine Learning in Medical Imaging
- Generative Adversarial Networks and Image Synthesis
- Face and Expression Recognition
- Advanced Malware Detection Techniques
- Neural Networks and Applications
- Digital Imaging for Blood Diseases
Stratford University
2024
Peking University Shenzhen Hospital
2022-2024
Hong Kong University of Science and Technology
2024
University of Hong Kong - Shenzhen Hospital
2024
Stanford University
2023-2024
Fudan University
2016-2021
BACKGROUND. The tumor immune microenvironment can provide prognostic and therapeutic information. We aimed to develop noninvasive imaging biomarkers from computed tomography (CT) for comprehensive evaluation of context, investigate their associations with prognosis immunotherapy response in gastric cancer (GC).
As a widely used imaging modality in the medical field, ultrasound has been applied community medicine, rural and even telemedicine recent years. Therefore, development of portable devices become popular research topic. However, limited size usually degrades quality, which reduces diagnostic reliability. To overcome hardware limitations improve image quality devices, we propose novel generative adversarial network (GAN) model to achieve mapping between low-quality images corresponding...
In recent years, plane-wave imaging (PWI) has attracted considerable attention because of its high temporal resolution. However, the low spatial resolution PWI limits clinical applications, which inspired various studies on reconstruction PW ultrasound images. Although compounding methods and traditional approaches can improve image quality, these techniques decrease Since learning fully reserve ultrasounds, a novel convolutional neural network (CNN) model for spatial-temporal images is...
Deep learning-based diagnosis is becoming an indispensable part of modern healthcare. For high-performance diagnosis, the optimal design deep neural networks (DNNs) a prerequisite. Despite its success in image analysis, existing supervised DNNs based on convolutional layers often suffer from their rudimentary feature exploration ability caused by limited receptive field and biased extraction conventional (CNNs), which compromises network performance. Here, we propose novel named manifold...
Symbolic regression (SR) has emerged as a pivotal technique for uncovering the intrinsic information within data and enhancing interpretability of AI models. However, current state-of-the-art (sota) SR methods struggle to perform correct recovery symbolic expressions from high-noise data. To address this issue, we introduce novel noise-resilient (NRSR) method capable recovering Our leverages reinforcement learning (RL) approach in conjunction with designed gating module (NGM) learn selection...
Abstract Predicting the occurrence of nonproliferative diabetic retinopathy (NPDR) using biochemical parameters is invasive, which limits large‐scale clinical application. Noninvasive retinal oxygen metabolism and hemodynamics 215 eyes from 73 age‐matched healthy subjects, 90 patients without DR, 40 NPDR, 12 DR with postpanretinal photocoagulation were measured a custom‐built multimodal imaging device. Diabetic underwent examinations. Two logistic regression models developed to predict NPDR...
In the medical ultrasound field, ultrafast imaging has recently become a hot topic. However, diagnostic reliability of high-frame rate plane-wave (PW) is reduced by its low-quality images. The equipment on market usually adopts line-scanning mode, which can obtain high-quality images at very low frame rate. addition, many proven data-driven image processing methods are trained line-scan Since gray-level distributions and PW different, these distribution-sensitive cannot be generalized to...
Recently, advances in computers and high-speed communication tools have led to enhancements remote medical consultation research.Laws some localities require hospitals encrypt patient information (including images of the patient) before transferring data over a network.Therefore, developing suitable encryption algorithms is quite important for modern medicine.This paper demonstrates digital image algorithm based on chaotic mapping, which uses no-period no-convergence properties sequence...
Introduction Age-related macular degeneration (AMD) is a prevalent, chronic and progressive retinal degenerative disease characterized by an inflammatory response mediated activated microglia accumulating in the retina. In this study, we demonstrate therapeutically effects underlying mechanisms of microglial repopulation laser-induced choroidal neovascularization (CNV) model exudative AMD. Methods The CSF1R inhibitor PLX3397 was used to establish treatment paradigm for Neovascular leakage...
Background: Tertiary lymphoid structures (TLSs) are associated with favorable prognosis and enhanced response to anti-cancer therapy. A digital assessment of TLSs could provide an objective alternative that mitigates variability inherent in manual evaluation. This study aimed develop validate a gene panel based on biological prior knowledge for TLSs, further investigate its associations survival multiple therapies. Materials Methods: The present involved 1,704 patients gastric cancer from...
Recently, the use of portable equipment has attracted much attention in medical ultrasound field. Handheld devices have great potential for improving convenience diagnosis, but noise-induced artifacts and low resolution limit their application. To enhance video quality handheld devices, we propose a low-rank representation multipathway generative adversarial network (LRR MPGAN) with cascade training strategy. This method can directly generate sequential, high-quality clear tissue structures...
The retina is one of the most metabolically active tissues in body. dysfunction oxygen kinetics closely related to disease and has important clinical value. Dynamic imaging comprehensive analyses depend on fusion structural functional high spatiotemporal resolution. But it's currently not clinically available, particularly via a single device. Therefore, this work aims develop retinal analysis (ROKIA) technology by integrating dual-wavelength with laser speckle contrast modalities, which...
A multi-modality image fusion can process images of certain organs or issues which were collected from diverse medical imaging equipment. The extract complementary information and integrate into with more comprehensive information. provide that was combined anatomical physiological for doctors bring convenience diagnosis. Basically, the thesis mainly studies MRI CT images, while taking cerebral infraction-suffered patients’ as example. Furthermore, T1 DWI sequences are respectively carrying...
To investigate changes in foveal avascular area (FAZ) and retinal vein diameter patients with occlusion (RVO) after intravitreal ranibizumab, to analyze the correlation between ranibizumab therapy visual gain. This retrospective study enrolled 95 eyes of who had accepted three consecutive monthly injections, including 50 branch RVOs (BRVOs) 45 central (CRVOs). BRVOs were divided into ischemia group (n = 32) non-ischemia 18), CRVOs also 28) 17). Comprehensive ophthalmic examinations performed...
Embedding high-dimensional data onto a low-dimensional manifold is of both theoretical and practical value. In this paper, we propose to combine deep neural networks (DNN) with mathematics-guided embedding rules for embedding. We introduce generic network (DEN) framework, which able learn parametric mapping from space space, guided by well-established objectives such as Kullback-Leibler (KL) divergence minimization. further recursive strategy, called (DRE), make use the latent...
Adequate lesion detection is critical in cystoscopy to improve bladder cancer resection and decrease recurrence. In particular, flat-appearing such as carcinoma situ difficult discern by standard white-light (WLC). The adoption of blue-light (BLC), an adjunct imaging technique, remains modest due the expensive equipment required. We developed a deep-learning algorithm, CystoNet-F, for augmented flat lesions on WLC. CystoNet-F was designed augment WLC incorporating domain translation...
Abstract Purpose: This study introduces a novel multimodal-attention-based virtual mIF staining (MAS) system designed for efficient, reliable from label-free fluorescence images. Our goal was to overcome the performance bottleneck and time-cost limitations associated with traditional techniques thereby enhance their clinical utility. Method Materials: approach involved developing cutting-edge MAS model. model is built upon sophisticated end-to-end generative convolutional neural network...