China 
Africa 
Explore chapters and articles related to this topic
Advances in Artificial Intelligence Applied to Heart Failure
Published in Kayvan Najarian, Delaram Kahrobaei, Enrique Domínguez, Reza Soroushmehr, Artificial Intelligence in Healthcare and Medicine, 2022
Jose M. García-Pinilla, Francisco Lopez Valverde
Adler et al. used an ML algorithm (by training a boosted decision tree algorithm to relate a subset of the patient data with a very high or very low mortality risk) to capture correlations between patient characteristics and mortality. 5,822 hospitalized and ambulatory patients with HF were included. From this model a risk score was derived; and it was able to discriminate between low and high risk of death by identifying eight variables (diastolic blood pressure, creatinine, blood urea nitrogen, hemoglobin, white blood cell count, platelets, albumin, and red blood cell distribution width) (Adler et al., 2020).
Compositional isotemporal substitution analysis of physical activity, sedentary behaviour and cardiometabolic biomarkers in US adults: A nationally representative study
Published in European Journal of Sport Science, 2023
Jinqun Cheng, Yanhong Huang, Zhiqiang Ren, Peng Xu, Jianyi Tan, Baoying Huang, Yue Chen, Ziqiang Lin, Yanhui Gao
Cardiometabolic biomarkers were measured by trained technicians at the mobile examination centre according to standard procedures. In this study, cardiometabolic biomarkers were divided into the following categories: 1) markers of inflammation, including CRP, WBCs, segmented neutrophils and red blood cell distribution width (RDW); 2) markers of lipid metabolism, including high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), total cholesterol and TG; 3) markers of glucose metabolism, including insulin, insulin resistance homeostasis model assessment (HOMA-IR), blood glycohemoglobin (HbA1c), fasting glucose, and 2-h glucose; and 4) blood pressure, systolic blood pressure and diastolic blood pressure. Among them, LDL-C, TG, insulin, HOMA-IR, fasting glucose, and 2-h glucose were fasting biomarkers. HOMA-IR was calculated as: fasting glucose (mmol/L) * fasting insulin (pmol/L) /22.5. Two-hour glucose levels were obtained from fasting participants in the 2005–2006 survey only, via oral glucose tolerance test (OGTT). Before analyzing each cardiometabolic biomarker, participants who lacked the corresponding biomarker information were excluded. Detailed descriptions of the procedures and methods used to measure each cardiometabolic biomarker are available on the NHANES website.
Acute and two-week inhalation toxicity studies in rats for Polyalphaolefin (PAO) fluid
Published in Journal of Toxicology and Environmental Health, Part A, 2021
David R. Mattie, Matthew D. Wegner, Brian A. Wong, R. Arden James, Karen L. Mumy, Shawn M. McInturf, Barry J. Marcel, Teresa R. Sterner
Blood samples were only taken from the 2-week recovery cohort animals at the time of necropsy to look for possible delayed effects. Samples of whole blood with anticoagulant were evaluated using a blood analyzer (Hemavet 950, Drew Scientific, Dallas, TX) for the following metrics: neutrophils, lymphocytes, monocytes, eosinophils, basophils, red blood cells, hemoglobin, hematocrit, mean corpuscular volume, mean corpuscular hemoglobin, mean corpuscular hemoglobin concentration, red blood cell distribution width, platelets, and mean platelet volume. Plasma samples were examined using a chemistry analyzer (Vet Test 8008 and Vet Lyte, IDEXX Laboratories, Westbrook, ME) for albumin, alkaline phosphatase, alanine transaminase, aspartate aminotransferase, blood urea nitrogen, cholesterol, creatine kinase, creatinine, globulins, glucose, total bilirubin, total protein, triglycerides, sodium, potassium, and chloride.
Robust logistic regression tree for subgroup identification in healthcare outcome modeling
Published in IISE Transactions on Healthcare Systems Engineering, 2020
When a single predictor for the outcome of interest is not specified, we first identify the important variables that affect the outcome among the pool of covariates. Since many covariates in this dataset are categorical variables, group LASSO is a suitable variable selection method. Seven variables are selected: ER visits (the number of emergency room visits), Hospitalization (the number of hospital stays), RDW (red blood cell distribution width), Age, LAMA (whether the patient was treated with long-acting muscarinic antagonists), Alcohol (whether or not the patient uses alcohol), and Antibiotic (whether the patient was treated with antibiotics). The predominant one, ER visits, is used as the predictor for logistic regression at each node, and other variables serve as split variables in the tree. In constructing the tree, the significance level for the parameter instability test is set to 0.05, the minimum sample size per node is set to 150, and the parameter settings of the proposed method are the same as in Section 5.