A5101981730- Radiomics and Machine Learning in Medical Imaging
- Web Data Mining and Analysis
- COVID-19 diagnosis using AI
- Health disparities and outcomes
- Advanced Text Analysis Techniques
- Text and Document Classification Technologies
- AI in cancer detection
- biodegradable polymer synthesis and properties
- Rough Sets and Fuzzy Logic
- Lung Cancer Diagnosis and Treatment
- Higher Education and Teaching Methods
- Polymer crystallization and properties
- Image and Signal Denoising Methods
- Data Mining Algorithms and Applications
- Advanced Image Fusion Techniques
- Resilience and Mental Health
- Water-Energy-Food Nexus Studies
- Biometric Identification and Security
- Aging and Gerontology Research
- Microplastics and Plastic Pollution
- Soil Mechanics and Vehicle Dynamics
- Advanced Computational Techniques and Applications
- Medical Imaging and Analysis
- Education and Work Dynamics
- Data Management and Algorithms
Changchun University of Science and Technology
2019-2024
China Three Gorges University
2024
State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering
2024
Nanyang Normal University
2024
South China Agricultural University
2023
Key Laboratory of Guangdong Province
2023
Biosensores (Spain)
2022
Shandong Management University
2020-2022
Xihua University
2022
BioMimetic Systems (United States)
2022
Medical image quality is highly relative to clinical diagnosis and treatment, leading a popular research topic of medical denoising. Image denoising based on deep learning methods has attracted considerable attention owing its excellent ability automatic feature extraction. Most existing for adapted certain types noise have difficulties in handling spatially varying noise; meanwhile, detail losses structure changes occurred the denoised image. Considering context perception preserving, this...
Partial slow-release fertiliser substitution for urea combined with water-saving irrigation may synergistically improve rice yield, quality, water, and nitrogen (N) utilisation. A field experiment to evaluate different combinations of regimes: alternate wetting drying (AWD) flooding (FI), N strategies: N0 (no fertiliser); N1 (100% conventional N2 SCF – sulphur-coated N3 (70% + 30% urea), N4 (50% 50% urea) on efficient production. Results indicated that higher rates (N2 N3) increased total...
Medical image segmentation constitutes a crucial step in the analysis of medical images, possessing extensive applications and research significance within realm practice. Convolutional neural network achieved great success segmentation. However, acquiring large labeled datasets remains unattainable due to substantial expertise time required for labeling, as well heightened patient privacy concerns. To solve scarce data, we propose powerful Domain Tuning SAM images (DT-SAM). We construct an...
<abstract> <p>An important prerequisite for improving the reliability of lung cancer surveillance and clinical interventions is accurate nodule segmentation. Although deep learning effective at performing medical image segmentation, CT heterogeneity, size, shape, location variations, convolutional localized feature extraction characteristics, receptive field limitations continuous downsampling, lesion edge information losses, fuzzy boundary segmentation challenges, low accuracy...
Abstract N , ′‐Ethylenebis(12‐hydroxystearamide) (EBH), an aliphatic diamide organic compound with environmental friendliness, was selected as effective nucleating agent to be incorporated into poly(1,4‐butylene adipate) (PBA) polyester. EBH increased the crystallization temperature, crystallinity, crystal density and thermal stability, shortened time of PBA. slowed down critical temperature PBA α ‐crystal, accelerated β ‐to‐ phase transition rate tailored polymorphic crystalline structure...
With the development of robot-assisted minimally invasive surgery, enhanced automatic context recognition surgical procedures is becoming critical to improve surgeon performance and patient safety. Deep neural networks can be efficiency at identifying phases analyzing procedures. However, hard-to-identify frames in videos tend reduce accuracy recognition. This research adds SE attention mechanism ResNeXt101 extract image features recognize phases, while using Online Hard Frame Mapper (OHFM)...
Text classification is the key technology for topic tracking, and vector space model (VSM) one of most simple effective topics representation. Feature selection algorithm in VSM an important means data pre-processing, it can reduce dimension improve generalization ability algorithm. Therefore, necessary feature algorithms to be in-depth extensive research. So we study how affect tracking. Then get variation law that they add up their optimal values Finally, TDT evaluation methods prove...
The hard bottom layer of a paddy field has great influence on the driving stability and operation quality efficiency intelligent farm machinery. For machinery, continuous improvements in accuracy unmanned precision operations are needed to realize rice farming. In context machinery operation, complicated situation makes it difficult quantify local characteristics fields. this paper, an direct seeding machine chassis is used maneuver collect information simultaneously. This design data...
Abstract The aim of the current study was to explore delayed retirement intention and its influencing factors in a large sample older Chinese workers. A cross‐sectional questionnaire survey conducted among convenience 2776 Multinomial logistic regression analyses were performed separately for male female workers examine association between work characteristics, family factors, personal health neuroticism trait intention. Results showed that had low intention, while more reluctant delay than...
Quantitative structure-activity relationships (QSAR) were developed to predict the inhibition ratio of pyrrolidine derivatives on matrix metalloproteinase via heuristic method (HM) and gene expression programming (GEP). The descriptors 33 calculated by software CODESSA, which can calculate quantum chemical, topological, geometrical, constitutional, electrostatic descriptors. HM was also used for preselection 5 appropriate molecular Linear nonlinear QSAR models based GEP separately two...
In clinical diagnosis, automatic fracture detection can reduce misdiagnosis which caused by fatigue and inexperience of radiologists, provide support for reducing patient suffering preventing disease progression. This paper proposes a SK-DenseNet model to detect femur based on DenseNet network. It uses SK module adjust the size receptive fields adaptively, adopts Focal loss function update parameter, Grad-CAM method visualize femoral results. improve interpretability model. used dataset...
Shared mobility is an important part of a smart city transportation system. However, during the short period it has been in effect, privacy leakages have frequently occurred, and as travellers are increasingly paying attention to their privacy, hinder rapid development shared mobility. Therefore, explore origin paradox context propose some targeted measures for improvement. The attested numerous studies, where, despite obvious concern that will be compromised, users continued adopt services...