A5038537255- Radiomics and Machine Learning in Medical Imaging
- Lung Cancer Diagnosis and Treatment
- Advanced Sensor and Energy Harvesting Materials
- Tactile and Sensory Interactions
- Sarcoma Diagnosis and Treatment
- Advanced Image Processing Techniques
- Conducting polymers and applications
- Trauma Management and Diagnosis
- Total Knee Arthroplasty Outcomes
- Advanced MRI Techniques and Applications
- Electromagnetic wave absorption materials
- Non-Invasive Vital Sign Monitoring
- Radiation Dose and Imaging
- Pelvic and Acetabular Injuries
- Neurofibromatosis and Schwannoma Cases
- Salivary Gland Tumors Diagnosis and Treatment
- Advanced X-ray and CT Imaging
- Venous Thromboembolism Diagnosis and Management
- Ultrasound in Clinical Applications
- Medical Imaging Techniques and Applications
- Cardiac Imaging and Diagnostics
- Bone Tumor Diagnosis and Treatments
- Image Processing Techniques and Applications
- Advanced Antenna and Metasurface Technologies
- Image and Signal Denoising Methods
Sichuan University
2024-2025
Shanghai Changzheng Hospital
2021-2025
Second Military Medical University
2021
Abstract To develop a deep learning-based model for detecting rib fractures on chest X-Ray and to evaluate its performance based multicenter study. Chest digital radiography (DR) images from 18,631 subjects were used the training, testing, validation of learning fracture detection model. We first built pretrained model, simple framework contrastive visual representations (simCLR), using with training set. Then, simCLR was as backbone fully convolutional one-stage (FCOS) objective network...
<title>Abstract</title> Soft fluidic actuators and robotics have been widely developed due to their excellent flexibility strong load capacity. However, stretchable pumps, which serves as a core element produce pressurized energy for the soft integrated systems, remain restricted in mechanical compliance output pressures. Here, we report electrohydrodynamic pumps based on solid-state ionic conductors, contribute new charge injection mechanism significantly improve fluid pumping. The pump...
Computed tomography (CT) plays an important role in the diagnosis of lung nodules and early screening cancer. The purpose this study was to compare efficacy 1,024×1,024 matrix 512×512 artificial intelligence-based computer-aided (AI-CAD) for evaluating based on CT images. This retrospective analysis included 344 patients from two hospitals between January 2020 November 2023. images presenting smaller than 30 mm were reconstructed using matrix. We evaluated image quality AI-CAD detection...
Background Spread through air spaces (STAS) in lung adenocarcinoma means different treatment and worse prognosis. Purpose To construct a radiomics model based on CT scans to predict the presence of STAS stage IA adenocarcinoma, compared with traditional clinical model. Material Methods The study included 317 patients (median age = 57.21 years; range 45.84–68.61 years) pathologically confirmed adenocarcinoma. In total, 122 (38.5%) were diagnosed by pathology after operation. Two experienced...
Abstract Image quality control (QC) is crucial for the accurate diagnosis of knee diseases using radiographs. However, manual QC process subjective, labor intensive, and time-consuming. In this study, we aimed to develop an artificial intelligence (AI) model automate procedure typically performed by clinicians. We proposed AI-based fully automatic radiographs high-resolution net (HR-Net) identify predefined key points in images. then geometric calculations transform identified into three...
Triboelectric nanogenerators (TENGs) are emerging as new technologies to harvest electrical power from mechanical energy. With the distinctive working mechanism of triboelectric nanogenerators, they attract particular interest in healthcare monitoring, wearable electronics, and deformable energy harvesting, which raises requirement for highly conformable devices with substantial outputs. Here, a simple, low‐cost strategy fabricating stretchable ultra‐high output is developed. The TENG...
Electronic textiles with remarkable breathability, lightweight, and comfort hold great potential in wearable technologies smart human-machine interfaces. Ionic capacitive sensors, leveraging the advantages of electric double layer, offer higher sensitivity compared to traditional sensors. Current research on ion-capacitive sensors has focused mainly two-dimensional (2D) or three-dimensional (3D) device architectures, which show substantial challenges for direct integration compromise their...
•We employed a cascading approach of two 2D networks to achieve the functionality 3D image pulmonary nodules super-resolution target reconstruction CT among three axes.•We used pre-trained model for mapping from normal-resolution images high-resolution images, which enhanced stability and convergence speed results.•In addition evaluation we also relevant diagnostic indicators, makes our results more valuable clinical application. Rationale ObjectivesTo develop Dual...
The peri-tumor microenvironment plays an important role in the occurrence, growth and metastasis of cancer. aim this study is to explore value application a CT image-based deep learning model tumors peri-tumors predicting invasiveness ground-glass nodules (GGNs).
A high false-positive rate remains a technical glitch hindering the broad spectrum of application deep-learning-based diagnostic tools in routine radiological practice from assisting diagnosing rib fractures.To examine performance two versions software aiding radiologists fractures on chest computed tomography (CT) images.In total, 123 patients (708 fractures) were included this retrospective study. Two groups with different experience levels retrospectively reviewed images for concurrent...
Ultra-high resolution computed tomography (UHRCT) has shown great potential for the detection of pulmonary diseases. However, UHRCT scanning generally induces increases in time and radiation exposure. Super is a gradually prosperous application CT imaging despite higher dose. Recent works have proved that convolution neural network especially generative adversarial (GAN) based model could generate high-resolution using phantom images or simulated low data without extra Research used clinical...
To establish a radiomics nomogram based on MRI features combined with clinical characteristics for distinguishing pleomorphic adenoma (PA) from warthin tumor (WT). 294 patients PA (n = 159) and WT 135) confirmed by histopathology were included in this study between July 2017 June 2023. Clinical factors including data analyzed to model. 10 extracted selected T1WI FS-T2WI, used model calculate scores (Rad-scores). Rad-scores serve as crucial parameters Through Receiver operator (ROC) curve...
Triboelectric nanogenerators (TENGs) are emerging as new technologies to harvest electrical power from mechanical energy. With the distinctive working mechanism of triboelectric nanogenerators, they attract particular interest in healthcare monitoring, wearable electronics, and deformable energy harvesting, which raise requirement for highly conformable devices with substantial outputs. Here, a simple, low-cost strategy fabricating stretchable ultra-high output is developed. The TENG...
The incidence of total knee replacement (TKR) surgeries has increased, partly attributed to healthcare policies that cause premature and potentially unwarranted interventions. This raised concerns regarding a potential trend excessive surgeries.
Background Differentiating diagnosis between the benign schwannoma and malignant counterparts merely by neuroimaging is not always clear remains still confounding in many cases because of atypical imaging presentation encountered clinic lack specific diagnostic markers. Purpose To construct validate a novel deep learning model based on multi-source magnetic resonance (MRI) automatically differentiating spinal from benign. Material Methods We retrospectively reviewed MRI data 119 patients...