A5101517100- Coronary Interventions and Diagnostics
- Radiomics and Machine Learning in Medical Imaging
- Video Surveillance and Tracking Methods
- Medical Image Segmentation Techniques
- Image Retrieval and Classification Techniques
- Advanced Neural Network Applications
- Cardiac Imaging and Diagnostics
- Cerebrovascular and Carotid Artery Diseases
- Advanced Text Analysis Techniques
- Topic Modeling
- Advanced MRI Techniques and Applications
- RNA Interference and Gene Delivery
- Cardiac Valve Diseases and Treatments
- Cardiovascular Disease and Adiposity
- Stock Market Forecasting Methods
- Medical Imaging Techniques and Applications
- Recommender Systems and Techniques
- Data Quality and Management
- Human Pose and Action Recognition
- Tracheal and airway disorders
- Traditional Chinese Medicine Analysis
- Intracerebral and Subarachnoid Hemorrhage Research
- Cell death mechanisms and regulation
- Retinal Imaging and Analysis
- Advanced Image and Video Retrieval Techniques
Beijing University of Technology
2020-2025
China International Science and Technology Cooperation
2024
Northeastern University
2024
Hudson Institute
2023
Tencent (China)
2023
John Wiley & Sons (United Kingdom)
2023
John Wiley & Sons (United States)
2023
University of Oxford
2023
Central Hospital of Wuhan
2022
Beijing University of Posts and Telecommunications
2015-2022
Background: In recent years, the use of deep learning has become more commonplace in biomedical field and its development will greatly assist clinical imaging data interpretation. Most existing machine methods for coronary angiography analysis are limited to a single aspect. Aims: We aimed achieve an automatic multimodal recognise quantify angiography, integrating multiple aspects, including identification artery segments recognition lesion morphology. Methods: A set 20,612 angiograms was...
Predicting user responses, such as click-through rate and conversion rate, are critical in many web applications including search, personalised recommendation, online advertising. Different from continuous raw features that we usually found the image audio domains, input space always of multi-field mostly discrete categorical while their dependencies little known. Major response prediction models have to either limit themselves linear or require manually building up high-order combination...
Vaakyarthajnaana Hetu is an important concept in Indian literary and philosophical traditions that focuses on understanding the meaning of a sentence or text. It helps readers grasp connection between words, their context, intended message. This offers clear structured way to interpret texts, ensuring deeper complex ideas literature philosophy. In research, especially useful for studying classical texts. By using this approach, researchers can uncover author’s intention, resolve unclear...
Magnetic resonance imaging (MRI) has been one of the most powerful and valuable methods for medical diagnosis staging disease. Due to long scan time MRI acquisition, k-space under-samplings is required during acquisition processing. Thus, reconstruction, which transfers undersampled data high-quality magnetic imaging, becomes an important meaningful task. There have many explorations on interpolation reconstruction. However, these ignore strong correlation between target slice its adjacent...
The reproducibility of positron emission tomography (PET) radiomics features is affected by several factors, such as scanning equipment, drug metabolism time and reconstruction algorithm. We aimed to explore the role 3D local binary pattern (LBP)-based texture in increasing accuracy PET for predicting pelvic lymph node metastasis (PLNM) patients with cervical cancer. retrospectively analysed data from 177 squamous cell carcinoma. They underwent 18F-fluorodeoxyglucose (18F-FDG)whole-body...
To explore the effects and safety of low dose esketamine combined with propofol in elderly patients undergoing fibronchoscopy. Eighty who underwent painless fibronchoscopy our hospital from June 2021 to September were recruited,and randomly divided into experimental group (esketamine 0.15mg/ kg + 1mg/ kg) control (sufentanil 0.1 μg/ kg), 40 cases each group. There significant differences MAP, HR SpO2 T2, T3 T4 between groups (P < .05). Besides, there significantly on trend change 2 groups, a...
Traditional enhanced external counterpulsation (EECP) used for the clinical treatment of patients with coronary heart disease only assesses diastolic/systolic blood pressure (Q = D/S &gt; 1.2). However, improvement hemodynamic environment surrounding vascular endothelial cells arteries after long-term application EECP is basis treatment. Currently, quantitative mechanism not well understood. In this study, a standard 0D/3D geometric multi-scale model artery was established to simulate...
The scale of training data is significant in segmentation task, especially segmenting the medical coronary artery angiograms. Traditional semantic networks have been restricted this field, due to particularity cardiac angiography data, that is, it very difficult balance manual labeling costs and network accuracy. On basis these observations, we propose a new method generate so-called 'pseudo-precise' label complementary pipeline, which can improve performance on premise reducing labor as...
Convolutional neural networks (CNNs) have been widely used in medical image segmentation. Vessel segmentation coronary angiography remains a challenging task. It is great challenge to extract fine features of artery for due the poor opacification, numerous overlap different segments and high similarity between soft tissues an image, which results sub-optimal performance. In this paper, we propose adapted generative adversarial (GANs) complete conversion from semantic image. We implemented...
In the field of cardiac arterial interventional therapy, coronary angiography imaging provides key information to physicians for treatment strategy selection, while lesion identification process is time-consuming and error-prone even experienced doctors. This paper proposes a method automatic detection in based on deep learning convolution neural network very first time. We used 2925 medical images building model. Several lesions exist vessel each image. will regard these areas as objects...
Differences in concentration of molecules can cause different molecular diffusion. This issue has not been well studied the vascular remodeling process with regards to is-stent restenosis. study designed and built a model explore effect differences cell on remodeling. Finite element analysis (FEA) models agent-based (ABMs) were established simulate damage proliferation smooth muscle cells (VSMCs) caused by coronary artery stent implantation. The FEA simulated expansion artery, tensile stress...
The quarterly financial statement, or Form 10-Q, is one of the most frequently required filings for US public companies to disclose and other important business information. Due massive volume 10-Q enormous variations in reporting format, it has been a long-standing challenge retrieve item-specific information from that lack machine-readable hierarchy. This paper presents solution itemizing files by complementing rule-based algorithm with Convolutional Neural Network (CNN) image classifier....
Background: The conventional FFRct numerical calculation method uses a model with multi-scale geometry based upon CFD, and rigid walls. Therefore, important interactions between the elastic vessel wall blood flow are not routinely considered. Changes in resistance of coronary microcirculation during hyperaemia likewise typically incorporated using fluid-structure interaction (FSI) algorithm. It is likely that both have resulted errors. Objective: In this study we influence vascular...
Percutaneous coronary intervention with stent implantation is one of the most commonly used approaches to treat artery stenosis. Stent malapposition (SM) can increase incidence thrombosis, but quantitative association between SM distance and thrombosis poorly clarified. The objective this study determine biomechanical reaction mechanisms underlying induced by quantify effect different severity grades on thrombosis. thrombus simulation was performed in a continuous model based...
Cardiovascular disease (CVD) is one of the diseases with highest mortality rate in modern society, while chronic total occlusion (CTO) initial factor that influences success percutaneous coronary intervention (PCI), which most common treatments for CVD. In this work, novel deep convolutional neural networks (CNNs) are proposed to detect entry point CTO and classify its morphology according angiography automatically. Specifically, feature pyramid (FPN) module model fusion technique applied...
Standard methods of subspace clustering are based on self-expressiveness in the original data space, which states that a point can be expressed as linear combination other points. However, real raw form usually not well aligned with model. Therefore, it is crucial to obtain proper feature space for performing high quality clustering. Inspired by success Convolutional Neural Networks (CNN) extraction powerful features from visual and block diagonal prior learning good affinity matrix...