A5101651673- Remote Sensing and Land Use
- Image Processing Techniques and Applications
- Advanced Algorithms and Applications
- Tropical and Extratropical Cyclones Research
- Advanced Neural Network Applications
- Remote-Sensing Image Classification
- Anomaly Detection Techniques and Applications
- Advanced Vision and Imaging
- Adversarial Robustness in Machine Learning
- Bladder and Urothelial Cancer Treatments
- Advanced Sensor and Control Systems
- Radiomics and Machine Learning in Medical Imaging
- Environmental Changes in China
- Technology and Security Systems
- Ocular and Laser Science Research
- Time Series Analysis and Forecasting
- Soft Robotics and Applications
- Optical Coherence Tomography Applications
- Advancements in PLL and VCO Technologies
- Generative Adversarial Networks and Image Synthesis
- Data-Driven Disease Surveillance
- Machine Learning in Healthcare
- Bacillus and Francisella bacterial research
- Healthcare Technology and Patient Monitoring
- Robotic Mechanisms and Dynamics
Tianjin Medical University
2024-2025
Second Hospital of Tianjin Medical University
2024-2025
University of Chinese Academy of Sciences
2021-2024
Chongqing Metrology Quality Inspection and Research Institute
2022-2024
University of Wisconsin–Madison
2023
Hanshan Normal University
2021
Chongqing Institute of Green and Intelligent Technology
2021
Mudanjiang Normal University
2020
Nanjing University of Information Science and Technology
2014
Chinese Academy of Meteorological Sciences
2014
With convolution operations, Convolutional Neural Networks (CNNs) are good at extracting local features but experience difficulty to capture global representations. cascaded self-attention modules, vision transformers can long-distance feature dependencies unfortunately deteriorate details. In this paper, we propose a hybrid network structure, termed Conformer, take both advantages of operations and mechanisms for enhanced representation learning. Conformer roots in coupling CNN transformer...
Cancer-associated fibroblasts (CAFs) are found in primary and advanced tumours. They primarily involved tumour progression through complex mechanisms with other types of cells the microenvironment. However, essential fibroblasts-related genes (FRG) bladder cancer still need to be explored, there is a shortage an ideal predictive model or molecular subtype for immune therapeutic assessment cancer, especially muscular-invasive based on FRG.
As an antibody-drug conjugate (ADC), disitamab vedotin (RC48) is a promising treatment targeting ERBB2 for locally advanced and metastatic bladder cancer (BLCA). However, the subtype heterogeneity of muscle-invasive (MIBC) often leads to different therapeutic outcomes. In our study, we aim explore sensitivity differences mechanisms molecular subtypes MIBC RC48 develop strategy combination therapy against cancer. Using large-scale mRNA expression profile datasets, Western blotting,...
Abstract To evaluate model performance of simulating structural characteristics in boundary layer at the forested areas northern United States and compare thermal dynamic with five different parameterizations, high resolution mesoscale WRF was used this paper to conduct some experiments parameterization configurations. The results showed that schemes could successfully simulate strong turbulent mixing daytime inversion, inverse wet low level jet stable nighttime except for individual...
Microcomputer-based oscilloscope instruments have been developed recently because they can easily process, store, and transmit various measured signals with the help of computers. Although this type is workable available, at present time, we still do not know too much about quantitative relationships respect to system bandwidth, buffer size, channel numbers, loss waveform interval in data buffering mechanism an asynchronous microcomputer-based oscilloscope. We propose a modified queueing...
To transfer the representation capacity of large pre-trained models to lightweight models, knowledge distillation has been widely explored. However, conventional single-stage methods are prone getting stuck in task-specific knowledge, making it difficult retain task-agnostic which is crucial for model generalization. In this study, we propose generic-to-specific (G2SD), boost under assistance by masked image modeling. generic distillation, decoder a small encouraged align feature predictions...
Develop a series of quality control materials (QCMs) for ruby color grading. Ruby was synthesized using flame melting method, and the samples were cut polished to prepare same size glossiness. According GB/T 32863-2016, D50 standard light source is used, 9 experienced jewelry grading technicians collaborate determine value uncertainty. The determined values mainly include tone, brightness, saturation, with Munsell hue ranging from 2.78RP 1.25R, 3.4 6.7, chroma 7.4 11.0. This substances...
The task of anomaly detection is to separate anomalous data from normal in the dataset. Models such as deep convolutional autoencoder (CAE) network and supporting vector description (SVDD) model have been universally employed demonstrated significant success detecting anomalies. However, over-reconstruction ability CAE for can easily lead high false negative rate data. On other hand, SVDD has drawback feature collapse, which leads a decrease accuracy To address these problems, we propose...
Nature disasters play a key role in shaping human-urban infrastructure interactions. Effective and efficient response to natural is essential for building resilience sustainable urban environment. Two types of information are usually the most necessary difficult gather disaster response. The first about damage perception, which shows how badly people think that has been damaged. second geolocation awareness, means whereabouts made available. In this paper, we proposed novel mapping...
The occurrence of perioperative critical adverse events will affect the quality medical services and threaten safety patients. Using scientific methods to assess risk illness in period is great significance improve service ensure However, diagnosis treatment data patients are multi-source irregular, only one physiological information can not accurately reflect patient's condition. In previous studies, it found that a variety transmit human health or not, which be used evaluate physical...
Abstract There is little published evidence of predicting cancer genotypes directly from tissue histology, especially for breast and prostate cancers. Artificial intelligence (AI) enables discriminating extracting morphological features reflective the underlying genomic alterations at visual subvisual levels. We have built a morphology-based AI-powered platform genotyping, risk stratification outcome prediction that addresses needs treatment decision-making in cost effective timely fashion....
e17097 Background: About half of the prostate cancer (PCa) patients within an intermediate-risk group return to clinic with disease progression after first line therapy. While these can be stratified into Gleason Grades (GG) 3+4 and 4+3, studies indicate that high-risk still found in both groups. Using our AI platform, we have developed a morphometric biomarker by analyzing digitized H&E slide images (WSIs), which accurately predict early biochemical recurrence (within 36 months...
Objective To investigate the effects of endothelin-1 (ET-1) and endothelin-3 (ET-3) on expression transformation growth factor-beta 1 (TGF-β1) phosphorylation Smad 3 in malignant melanoma cell line, A375. Methods Cultured A375 cells were classified into 5 groups, i.e. control group (no stimulation), ET-1 (stimulated with ET-1), + BQ123 (treated BQ123), BQ788 BQ788), ET-3 ET-3), to receive different stimulation. The working concentrations 0, 0.1, 1, 10, 100 nmol/L for ET-3, 10 μmol/L BQ788....
With the popularization of intelligent electronic devices, mobile learning models have become an inevitable trend.By establishing platforms, universities can effectively make up for problems traditional methods, and reasonably introduce big data technology Internet to develop computers Applied teaching brings out advantages strengthens communication interaction with students, integrates classrooms innovate concepts continuously improve quality.