A5038783965- Radiomics and Machine Learning in Medical Imaging
- Medical Imaging Techniques and Applications
- Lung Cancer Diagnosis and Treatment
- Complex Network Analysis Techniques
- Text and Document Classification Technologies
- Bladder and Urothelial Cancer Treatments
- Spectroscopy and Chemometric Analyses
- Endometrial and Cervical Cancer Treatments
- Digital Media Forensic Detection
- Advanced X-ray and CT Imaging
- Colorectal Cancer Screening and Detection
- Urinary and Genital Oncology Studies
- Advanced Graph Neural Networks
- Advanced Steganography and Watermarking Techniques
- Advanced Neural Network Applications
- Topic Modeling
- Advancements in Battery Materials
- Image and Video Stabilization
- Infrared Target Detection Methodologies
- Advanced SAR Imaging Techniques
- Technology Adoption and User Behaviour
- Advanced Image Fusion Techniques
- Advanced Battery Materials and Technologies
- Irrigation Practices and Water Management
- Hydrology and Watershed Management Studies
Dongyang People's Hospital
2023-2024
Wenzhou Medical University
2023-2024
Sichuan University
2022-2023
PLA 306 Hospital
2019-2020
Anyang Institute of Technology
2020
Xihua University
2018-2020
Chinese Academy of Medical Sciences & Peking Union Medical College
2019
Peking Union Medical College Hospital
2019
Academia Sinica
2019
Academy of Medical Sciences
2019
Deep learning-based methods have achieved great success in target detection tasks of computer vision, but when it comes to Synthetic Aperture Radar (SAR) image ship detection, some new challenges appear because the wide swath images, diverse appearances ships and lack detail information, which make inefficient less effective. Aiming these issues, this paper, a lightweight feature optimizing network (LFO-Net) based on popular single shot detector (SSD) model is proposed for polarization SAR...
Maritime activities are essential aspects of human society. Accurate classification ships is vital for maritime surveillance and meaningful to numerous civil military applications. However, most studies conducted limited the coarse-grained ship classification. Few on fine-grained have been undertaken despite its accuracy practicability. In this study, we construct a new benchmark which consists 23 categories ships. Besides category label, contains several other attribute information. To...
Background The novel International Association for the Study of Lung Cancer (IASLC) grading system suggests that poorly differentiated invasive pulmonary adenocarcinoma (IPA) has a worse prognosis. Therefore, prediction IPA before treatment can provide an essential reference therapeutic modality and personalized follow-up strategy. This study intended to train nomogram based on CT intratumoral peritumoral radiomics features combined with clinical semantic features, which predicted was tested...
This study aimed to establish an effective model for preoperative prediction of tumor deposits (TDs) in patients with rectal cancer (RC).In 500 patients, radiomic features were extracted from magnetic resonance imaging (MRI) using modalities such as high-resolution T2-weighted (HRT2) and diffusion-weighted (DWI). Machine learning (ML)-based deep (DL)-based models developed integrated clinical characteristics TD prediction. The performance the was assessed area under curve (AUC) over...
In this study, we investigate the effects of flow and consumer attachment on e-loyalty SNS websites users. We build test a causal model using data obtained from Renren.com find support for all our hypotheses. The results indicate that plays an influential role attachment, also has direct effect referral loyalty online-stickiness e-loyalty. Consumer positively affects (online-stickiness loyalty), it fully mediates relationship between e- (both loyalty). implications future research directions...
In microblog networks, when a user posts microblog, other users may forward the post, and then forwarding process will bring about rapid dissemination diffusion of information. this paper, we propose comprehensive novel approach to predict behavior. Firstly, build feature sets that affect forwarding, such as interest topic,geographic location, aggregation coefficient,neighborhood overlap so on. These features are classified into four categories: characteristics, features, network structure...
The surgical approach and prognosis for invasive adenocarcinoma (IAC) minimally (MIA) of the lung differ. However, they both manifest as identical ground-glass nodules (GGNs) in computed tomography images, no effective method exists to discriminate them.We developed validated a three-dimensional (3D) deep transfer learning model IAC from MIA based on CT images GGNs. This uses 3D medical image pre-training (MedicalNet) fusion build classification network. Transfer was utilized end-to-end...
Two lithium ions can be reversibly embedded/detached from the structure of Li2FeSiO4, thus having a high theoretical discharge specific capacity. However, it has defects low conductivity and slow ion diffusion rate, which makes difficult to realize deintercalation more than one ion, its actual capacity is lower This greatly restricts development into commercial battery cathode materials. In this paper, Li2FeSiO4/Mgx Li2FeSiO4/Cux (x = 0.01, 0.02, 0.03, 0.05) are synthesized by solid-state...
In the intention of minimizing excessive X-ray radiation administration to patients, low-dose computed tomography (LDCT) has become a distinct trend in radiology. However, while lowering dose reduces risk patient, it also increases noise and artifacts, compromising image quality clinical diagnosis. most supervised learning methods, paired CT images are required, but such unlikely be available clinic. We present self-supervised model (Noise2Projection) that fully exploits raw projection...
Relation classification is a task that identifies entity's semantic relationships from unstructured text. Lately, the method based on convolution neural network (CNN) achieves competitive performance comparing with other complex-structured networks by only using standard layer, pooling layer and softmax layer. However, CNN difficultly extract effective features or even wrong when processing samples of large-pitch entities, decreasing their accuracy. Besides existing methods could obtain...
Micro-blog post retweet prediction is one of the key technologies for researching information dissemination. The factors that affect are multifaceted; However, existing methods mainly focus on research micro-blog attributes and network structure characteristics, do not fully consider interests users changing with time. To address this issue, we firstly propose some new measuring features, which user's influence, intimacy named entities users' intimacy. Meanwhile, in order to solve problem...
At present, the layout of sprinkler position in irrigation system mainly depends on manual design. When scale is large, difficult to realize, and unreasonable easy cause problems such as equipment water resources waste. In this paper, k-means algorithm machine learning applied agricultural irrigation, concept intelligent design proposed. Firstly, area simulated stored computer based inner outer point discrimination polygon, then clustered divided obtain each cluster. The central cluster best...
Upper tract urothelial carcinoma (UTUC) accounts for 5%–10% of all tumors. Radical nephroureterectomy is the standard treatment procedure. At present, different choices still exist treating ureteral end during laparoscopic bladder sleeve resection. Our center has adopted a new method end. This can increase operating space and reduce difficulty surgery compared with current methods.