A5051861692- Elevator Systems and Control
- Anesthesia and Pain Management
- Advanced Neural Network Applications
- Integrated Circuits and Semiconductor Failure Analysis
- Robotic Path Planning Algorithms
- Management of metastatic bone disease
- Surface Roughness and Optical Measurements
- Industrial Vision Systems and Defect Detection
- Lung Cancer Diagnosis and Treatment
- Machine Learning and ELM
- Image and Object Detection Techniques
- Ultrasound in Clinical Applications
- Mobile Agent-Based Network Management
- Hemodynamic Monitoring and Therapy
- Anesthesia and Sedative Agents
- Airway Management and Intubation Techniques
- Brain Tumor Detection and Classification
- Total Knee Arthroplasty Outcomes
- Respiratory Support and Mechanisms
- Power Line Inspection Robots
- Power Line Communications and Noise
- Nausea and vomiting management
- Radiation Dose and Imaging
- Energy Efficient Wireless Sensor Networks
- Reinforcement Learning in Robotics
Zhejiang University of Science and Technology
2022-2025
Army Medical University
2014-2019
Daping Hospital
2014-2019
Shanghai Jiao Tong University
2005
An approach to the problem of autonomous mobile robot obstacle avoidance using reinforcement learning neural network is proposed in this paper. Q-learning one kind method that similar dynamic programming and has a powerful ability store values. We integrate these two methods with aim ensure behavior complicated unpredictable environment. The simulation results show simulated can enhance its obviously finish given task complex
An improved YOLOv8n seamless fabric defect detection model is proposed to solve the current issues in defects factories this paper. The improvement paper first introduces SPPF_LSKA module, which not only optimizes extraction of multi-scale features but also enhances adaptability detecting different sizes by improving feature fusion mechanism, enabling efficient recognition both large-sized and small-sized defects. Secondly, CARAFE upsampling method used adaptively learn relationship between...
Convolutional neural network (CNN) is quite popular in computer vision, especially image classification with excellent performance. However, limited by the convolution kernels, CNN-based classifiers are hard to extract global feature from original image, while exact object locations environment included feature. One way improve extraction performance use graph (GNN) which can aggregate information through connection relationship of different nodes. In this work, a novel end-to-end...
We aim to assess efficacy and safety of remifentanil or sulfentanyl combined with propofol during painless gastroscopic examination in patients. In this study, 270 patients were randomly divided into 3 groups: was given only group P; PR; PS the examination. Efficiency P significantly higher than that PR (P<0.01) [corrected]. lower (P<0.05). Incidence chest wall rigidity oxygen desaturation (P<0.05), whereas there no difference between groups (P>0.05). Propofol could provide satisfying...
Lane detection is a key technical challenge in autonomous driving and advanced assistance systems. The current lane system has insufficient robustness low accuracy problem complex harsh environmental conditions, lacks research on visual cues without information. In order to accurately effectively detect multiple lanes challenging scenarios meet the real-time requirements of actual driving, method based recurrent cross attention mechanism proposed this paper. Firstly, by embedding module...
Objective To evaluate the predictive value of ultrasonography respiratory variability index inferior vena cava internal diameter (IVC-RVI) in preoperative fluid responsiveness evaluation hip fracture geriatric patients. Methods A retrospective case-control study was used to analyze 39 cases elderly fractures admitted Daping Hospital Army Medical Center from August 2017 February 2018, including 15 males and 24 females, aged 65-92 years [(80.9±7.9)years]. The heart rate (HR), mean...
Objective To evaluate the effect of multidisciplinary cooperative pain management on rapid recovery patients with total hip and knee arthroplasty. Methods A 120 patients, American Society Anesthesiologists physical status Ⅰ or Ⅱ, aged 20-64 yr, body mass index 18-25 kg/m2, were divided into 2 groups using a random number table method: test group (group T, n=66) control C, n=54). Multidisciplinary mode was adopted for in perioperative period while traditional used C. Numeric rating...