A5103216459- Advanced X-ray and CT Imaging
- Medical Imaging Techniques and Applications
- Cancer Immunotherapy and Biomarkers
- Radiation Dose and Imaging
- Lung Cancer Treatments and Mutations
- Medical Image Segmentation Techniques
- Epigenetics and DNA Methylation
- Radiomics and Machine Learning in Medical Imaging
- Advanced MRI Techniques and Applications
- Digital Radiography and Breast Imaging
- Lung Cancer Diagnosis and Treatment
- vaccines and immunoinformatics approaches
- MRI in cancer diagnosis
- Acute Ischemic Stroke Management
- Cancer Genomics and Diagnostics
- Advanced Radiotherapy Techniques
- Computational and Text Analysis Methods
- Cancer-related molecular mechanisms research
- Cancer, Lipids, and Metabolism
- Gene expression and cancer classification
- Immunotherapy and Immune Responses
- AI in cancer detection
- Gestational Trophoblastic Disease Studies
- Ferroptosis and cancer prognosis
- Computational Drug Discovery Methods
Zhujiang Hospital
2016-2022
Southern Medical University
2015-2022
The University of Texas Southwestern Medical Center
2015-2016
Southwestern Medical Center
2015
Research has shown that the lipid microenvironment surrounding colorectal cancer (CRC) is closely associated with occurrence, development, and metastasis of CRC. According to pathological images from National Center for Tumor diseases (NCT), University Medical Mannheim (UMM) database ImageNet data set, a model called VGG19 was pre-trained. A deep convolutional neural network (CNN), VGG19CRC, trained by migration learning method. VGG19CRC model, adipose tissue scores were calculated TCGA-CRC...
X-ray scatter is a significant barrier to image quality improvements in cone-beam computed tomography (CBCT). A moving blocker-based strategy was previously proposed simultaneously estimate and reconstruct the complete volume within field of view (FOV) from single CBCT scan. blocker consisting lead stripes inserted between source imaging object, moves back forth along rotation axis during gantry rotation. While promising results were obtained our previous studies, geometric design speed set...
Lung cancer, mainly lung adenocarcinoma, squamous cell carcinoma and small has the highest incidence cancer-related mortality worldwide. Platinum-based chemotherapy plays an important role in treatment of various cancer subtypes, but not all patients benefit from this regimen; thus, it is worth identifying who are resistant or sensitive to platinum-based therapy.The drug response sequencing data 170 lines were downloaded Genomics Drug Sensitivity Cancer (GDSC) database, support vector...
Computed tomography (CT) images will be severely damaged from low-mAs acquisition conditions. Seriously degraded CT may lead to diagnostic bias in clinics. It is vital assess image quality before diagnosis. However, lack of high-quality reference clinical practice makes the full-reference (FR)- and reduced-reference (RR)- assessment (IQA) models difficult for assessment. In addition, convolutional neural network (CNN) has been proven effective natural Meanwhile these CNN-based methods...
Purpose: In computed tomography perfusion (CTP) imaging, an initial phase CT acquired with a high‐dose protocol can be used to improve the image quality of later low‐dose protocol. For dynamic regions, signals in may not completely recovered if heavily regularizes iterative reconstruction process. The authors propose hybrid nonlocal means (hNLM) regularization model for CTP overcome limitation conventional prior‐image induced penalty. Methods: penalty was constructed by combining NLM...
Melanoma refers to a pigmented nevus with malignant changes. The preferred treatment for primary melanoma is surgical excision and postoperative radiotherapy, but the prognosis poor. Immune checkpoint inhibitors (ICIs) have been remarkably successful in different types of cancers, not all cancer patients can benefit from it. Therefore, it essential find predictable biomarkers improve accuracy treatment. In this study, we used survival analysis, gene panorama immune cell enrichment TMB GSEA...
Abstract: Peer feedback is a pedagogical strategy for peer learning. Despite recent indications of Large Language Models (LLMs) ' potential content analysis, there limited empirical exploration their application in supporting the process. This study enhances analytical approach to activities by employing state-of-the-art LLMs automated feature detection. research critically compares three models—GPT-3.5 Turbo, Gemini 1.0 Pro, and Claude 3 Sonnet— evaluate effectiveness The involved 69...
Purpose: In CT perfusion imaging, an initial phase acquired with a high‐dose protocol can be used to improve the image quality of later low‐dose protocol. For dynamic regions, signals in may not completely recovered if heavily regularizes iterative reconstruction process. To overcome limitation conventional prior induced penalty, we propose hybrid nonlocal means (NLM) regularization for CT. Methods: The penalty is constructed by combining NLM stationary region and region. regions are...
Urothelial cancer (UC) is one of the common refractory tumors and chemotherapy primary treatment for it. The advent immune checkpoint inhibitors (ICI) has facilitated development strategies UC patients. To screen out patients sensitive to ICI, researchers have proposed that PD-L1, tumor mutation burden TCGA molecular subtypes can be used as predictors ICI efficacy. However, performance these needs further validation. We need identify novel biomarkers ICI. In our study, we collected data two...
In the clinics, unenhanced CT (unCT) are first used for acute ischemic stroke detection, but sometimes it is difficult to delineate infarction in unCT few hours. Currently, deep learning methods can be segmentation, they based on a large amount of annotation data, especially imaging. Meanwhile, Diffusion-weighted magnetic resonance imaging (DWI) clearly distinguish infarcted brain tissue from healthy region. this work, we investigate whether external information, i.e., DWI images, help more...
Purpose: Low‐dose CBCT is desired in various clinical applications. Iterative image reconstruction algorithms have shown advantages suppressing noise low‐dose CBCT. However, due to the smoothness constraint enforced during process, edges may be blurred and features lose reconstructed image. In this work, we proposed a new penalty design preserve by iterative algorithms. Methods: minimizing penalized weighted least‐squares (PWLS) objective function. Binary Robust Independent Elementary...
High imaging dose has been a concern in computed tomography perfusion (CTP) as repeated scans are performed at the same location of patient. On other hand, signal changes only occur limited regions CT acquired different time points. In this work, we propose new reconstruction strategy by effectively utilizing initial phase high-quality to reconstruct later with low-dose protocol. proposed strategy, is considered base image and enhanced (ES) reconstructed directly minimizing penalized...
Energy-resolving computed tomography (ErCT) has the ability to acquire energy-dependent measurements simultaneously and quantitative material information with improved contrast-to-noise ratio. Meanwhile, ErCT imaging system is usually equipped an advanced photon counting detector, which expensive technically complex. Therefore, clinical scanners are not yet commercially available, they in various stage of completion. This makes researchers less accessible images. In this work, we investigate...
Abstract Background: Lung cancer, mainly including lung adenocarcinoma, squamous cell carcinoma and small is the cancer with highest incidence cancer-related mortality in world. Platinum-based chemotherapy plays an important role treatment of various subtypes, but not all patients can benefit from it, so it worth identifying who are resistant or insensitive. Method: The drug response sequencing data 170 lines were downloaded Genomics Drug Sensitivity Cancer (GDSC) database, support vector...