China
Africa
Explore chapters and articles related to this topic
Why One Needs Statistical Techniques
Published in David J. Smith, Sam Samuel, Basic Statistical Techniques for Medical and Other Professionals, 2021
Much of statistical analysis consists of drawing conclusions from a set of data or of comparing two or more sets of data. The data in question is nearly always a sample drawn from a wider (larger) population. It is therefore important to think about whether the data that one has gathered is, indeed, a representative sample of the population of interest.
Mathematics for medical imaging
Published in Ken Holmes, Marcus Elkington, Phil Harris, Clark's Essential Physics in Imaging for Radiographers, 2021
Statistical analysis may again be useful, and there are a number of common tests which may be used to analyse data:Analysis of variance (ANOVA)Chi-squared testMann–Whitney U testt-test
Characterization of Uncertainty
Published in Samuel C. Morris, Cancer Risk Assessment, 2020
The classical approach to defining ranges (or confidence limits) is statistical analysis. Harking back again to our lake, if we measured the temperature every day for a year, and found the daily temperature had a mean of 25 and a standard deviation of 5, we could calculate the 95% confidence range of the mean as
Lack or insufficient predialysis nephrology care worsens the outcomes in dialyzed patients – call for action
Published in Renal Failure, 2022
Andrzej Milkowski, Tomasz Prystacki, Wojciech Marcinkowski, Teresa Dryl-Rydzynska, Jacek Zawierucha, Jacek S. Malyszko, Pawel Zebrowski, Konrad Zuzda, Jolanta Małyszko
Data was collected for each patient prior to the first visit: vascular access (temporary CVC permanent catheter, AV fistula, and graft), age, BMI, systolic and diastolic arterial pressure, laboratory tests: eGFR, urea concentration, 50 Hb, phosphate, albumin, PTH, glucose, accompanying diseases: diabetes, hypertension, malignancy, Charlson comorbidity index, and dose of ESA determined in the first month of dialysis (darbopoetin doses converted to erythropoietin alfa). During 13-month observation the following data was collected: number of hospitalizations and length of stay, patients’ outcomes, i.e., survival, kidney transplantation, and mortality. The collected data was subjected to statistical analysis. Since only the available data from the EUCLiD system was analyzed, written consent was not needed in accordance with the regulations in force at Fresenius Nephrocare in Poland and Ethics Committee at the Warsaw Medical University.
Partial least squares regression with compositional response variables and covariates
Published in Journal of Applied Statistics, 2021
Jiajia Chen, Xiaoqin Zhang, Karel Hron
Regression analysis is one of the most commonly used multivariate statistical analysis methods. The study on linear regression analysis with random compositions is already quite mature, and focuses on three types: Type 1 refers to the model with compositional covariates and real response [2,7,16,20]; Type 2 relates to the model with compositional response and real covariates [10,14,28,29]; the response and covariates in Type 3 are both compositional [6,32]. But when the number of parts of all compositional covariates increases, the existing linear regression models with random compositions are not applicable, because the least squares method used in these models fails. In that case, partial least squares (PLS) regression can be used. The earlier research for Type 1 introduced the logcontrast PLS [15], later discriminant PLS analysis in clr coefficients or ilr coordinates was proposed [13,18]. Type 3 was studied by Wang et al. who proposed PLS regression in clr coefficients and the hierarchical PLS regression in clr coefficients, where there is one compositional response variable and one or more than one compositional covariate [31].
Molecular tissue profiling by MALDI imaging: recent progress and applications in cancer research
Published in Critical Reviews in Clinical Laboratory Sciences, 2021
Pey Yee Lee, Yeelon Yeoh, Nursyazwani Omar, Yuh-Fen Pung, Lay Cheng Lim, Teck Yew Low
Statistical analysis is then performed on the data using either unsupervised or supervised approaches [80]. Unsupervised methods such as hierarchical clustering analysis, principal component analysis, and t-distributed stochastic neighbor embedding perform clustering using unlabeled data as the input to uncover patterns based on data similarities for exploratory analysis [81]. Supervised approaches such as support vector machines and random decision forest perform classification on labeled data with known outcomes to identify new patterns and features from MALDI imaging datasets [68]. In the initial training phase, classifiers are built to discriminate the input data using different techniques such as support vector machines and random forests. In the subsequent validation phase, the performances of the classifiers in terms of parameters such as sensitivity, specificity, positive predictive value, and negative predictive value are evaluated using the same training set or an external dataset.