A5103224904- Machine Learning in Healthcare
- MicroRNA in disease regulation
- Neuroinflammation and Neurodegeneration Mechanisms
- Hydrogen's biological and therapeutic effects
- Cancer, Hypoxia, and Metabolism
- Lipid metabolism and disorders
- Artificial Intelligence in Healthcare and Education
- Cancer-related molecular mechanisms research
- Healthcare cost, quality, practices
- Transplantation: Methods and Outcomes
- Peroxisome Proliferator-Activated Receptors
- Sulfur Compounds in Biology
- Autophagy in Disease and Therapy
- Ethics in Clinical Research
- Sepsis Diagnosis and Treatment
- Privacy-Preserving Technologies in Data
- Vagus Nerve Stimulation Research
- Vitamin C and Antioxidants Research
- Artificial Intelligence in Healthcare
- Neurological Disease Mechanisms and Treatments
- Eicosanoids and Hypertension Pharmacology
- Employee Performance and Management
- Neural Networks Stability and Synchronization
- Work-Family Balance Challenges
- Advanced Causal Inference Techniques
Duke University
2021-2025
George Washington University
2020-2021
China Three Gorges University
2020
Tianjin Medical University General Hospital
2018-2019
Tianjin Hospital
2019
Tianjin Medical University
2018
Concord Repatriation General Hospital
2007
Royal Prince Alfred Hospital
2007
Abstract One of the key limitations electronic health records (EHR) data is that not all care encounters are observed. The degree to which patient information captured referred as observability. Poor observability, and in particular differential can lead biased estimates inference. As such, understanding observability important EHR based studies. In this study, we propose using external with known assess overall EHR. We also construct a test for target dataset. Using principles from...
We describe two immunocompetent patients with tuberculous cranial pachymeningitis. Both underwent biopsy after focal dural thickening was identified on MRI. Histopathologic examination of tissue revealed necrotizing granulomatous inflammation. PCR for Mycobacterium tuberculosis DNA negative CSF but positive tissue. responded to antituberculous therapy. Although uncommon as a cause pachymeningitis, should be considered, since it responds well treatment.
Electronic health records have incomplete capture of patient outcomes. We consider the case when observability is differential across a predictor. Including such predictor (sensitive variable) can lead to algorithmic bias, potentially exacerbating inequities.We define bias for clinical prediction model (CPM) as difference between true and estimated risk, that differs sensitive variable. illustrate genesis via 2-stage process, where conditional on having outcome interest, differentially...
Objective To evaluate the effect of dexmedetomidine on expression hypoxia-inducible factor-1α(HIF-1α)during endotoxin-caused apoptosis in macrophages mice. Methods Mouse macrophage cell line RAW264.7 cultured vitro were seeded 6-well or 96-well plates and divided into 4 groups(n=16 each)when confluence reached 60%-70% using a random number table method: control group(group Con), Dex), lipopolysaccharide(LPS)group, LPS plus LPS+ Dex). Phosphate buffer solution was added group...
Objective To investigate the changes in FUNDC1/microtubule-associated protein 1 light chain 3Ⅱ (LC3Ⅱ) signaling pathway during sepsis-induced liver injury mice. Methods Thirty-two clean-grade healthy male C57BL/6 mice, aged 6 weeks, weighing 20-25 g, were divided into sham operation group (n=8) and sepsis (n=24) using a random number table method.Sepsis was induced by cecal ligation puncture.Blood samples obtained at 24 h after 6, 12 establishing model for determination of...
Objective To evaluate the role of autophagy in dexmedetomidine-induced reduction lipopolysaccharide(LPS)-caused inflammatory responses macrophages mice. Methods Mouse macrophage cell line RAW264.7 cultured vitro were seeded 6-well or 96-well plates and divided into 4 groups (n=20 each) when confluence reached 60% using a random number table method: control group (group Con), LPS group, plus dexmedetomidine LPS+ DEX), inhibitor 3-MA DEX+ 3-MA). PBS was added cells for 12 h Con.LPS at...
Objective To evaluate the effect of hydrogen on blood brain barrier mice with sepsis-associated encephalopathy (SAE). Methods A total 100 adult male ICR mice, aged 6-8 weeks, weighing 20-25 g, were divided into 4 groups(n=25 each) using a random number table method: sham operation group (group Sham), plus Sham+ H), SAE and SAE+ H). Sepsis was induced by cecal ligation puncture (CLP). H groups inhaled 2% for 1 h starting from 6 after CLP, respectively.At 24 Evans blue (EB) injected via...
Objective: To investigate the endothelial protective effects of simvastatin on coagulation system in septic rats. Methods: A total 54 SD male rats were divided into 3 groups. Six healthy intraperitoneally injected with normal salineas control group. Twenty-four group saline followed by lipopolysaccharide 2.5 mg. Study had 24 lipopolysaccharide. Plasma von Willebrand factor (vWF), thrombomodulin (TM), platelet activating (PAF) and antithrombin-Ⅲ (AT-Ⅲ) tested at 1 h, 6 h 12 after treatment....
Abstract Motivation Traditional regression models are limited in outcome prediction due to their parametric nature. Current deep learning methods allow for various effects and interactions have shown improved performance, but they typically need be trained on a large amount of data obtain reliable results. Gene expression studies often small sample sizes high dimensional correlated predictors so that traditional not readily applicable. Results In this article, we proposed peel learning,...