PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Linear Models”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,207 records · Page 67Linked to original sources
PubMed1990

A two-phase linear regression model for biologic half-life data.

In estimating the biologic half-life of an infused drug or biologic agent, one very frequently used model is the biexponential, which reflects a two-compartment physiologic model. The difficulty in using this model is that it is nonlinear in the parameters and requires relatively sophisticated analysis. Furthermore, an interval estimate is not usually reported. We propose the use of a two-phase linear regression approach, which is tantamount to breaking down the model into two straight lines based on the selection of the logarithm of concentration as the ordinate and time as the abscissa. We show how to determine the joint or changeover point for the two lines using a simple iterative procedure, how to select between a one-phase and a two-phase model, and how to provide a simple confidence interval estimate for half-life when it exists. An example using data from a study of Factor VIII pharmacokinetics is given.

Factor VIII↗
PubMed2006

Reliability of linear measurements on a virtual bilateral cleft lip and palate model.

OBJECTIVE: To assess the reliability and validity of measurements performed on three-dimensional virtual models of neonatal bilateral cleft lip and palate patients, compared with measurements performed on plaster cast models. MATERIALS AND METHODS: Ten high-quality plaster cast models of bilateral cleft lip and palate patients were scanned with an LDI-scanner to obtain a three-dimensional virtual model. Linear measurements were performed on the plaster cast models using a digital caliper and also on the three-dimensional virtual model using Viscam RP version 2.1 software. The measurements were performed by two observers on two occasions. RESULTS: Intraclass correlations ranging from .81 to .96 were found for all measurements except the measurement between the constructed reference point pr and reference point i (intraclass correlation = .40). A post hoc procedure in which top-view screen prints of the three-dimensional virtual model were used to perform the measurement between reference points pr and i demonstrated an intraclass coefficient of .90. CONCLUSIONS: Three-dimensional virtual models obtained by laser scanning neonatal cast models of bilateral cleft lip and palate patients can be used reliably and validly to perform linear measurements between existing reference points on the surface of the model using Viscam RP version 2.1 software. Measurements between reference points constructed outside the surface of the model cannot be validly performed on the three-dimensional virtual model with the software used in this study. For these measurements, top-view screen prints of the three-dimensional virtual model can be used.

Cephalometry↗
PubMed2005

Accurate prediction of HIV-1 drug response from the reverse transcriptase and protease amino acid sequences using sparse models created by convex optimization.

MOTIVATION: Genotype-phenotype modeling problems are often overcomplete, or ill-posed, since the number of potential predictors-genes, proteins, mutations and their interactions-is large relative to the number of measured outcomes. Such datasets can still be used to train sparse parameter models that generalize accurately, by exerting a principle similar to Occam's Razor: When many possible theories can explain the observations, the most simple is most likely to be correct. We apply this philosophy to modeling the drug response of Type-1 Human Immunodeficiency Virus (HIV-1). Owing to the decreasing expense of genetic sequencing relative to in vitro phenotype testing, a statistical model that reliably predicts viral drug response from genetic data is an important tool in the selection of antiretroviral therapy (ART). The optimization techniques described will have application to many genotype-phenotype modeling problems for the purpose of enhancing clinical decisions. RESULTS: We describe two regression techniques for predicting viral phenotype in response to ART from genetic sequence data. Both techniques employ convex optimization for the continuous subset selection of a sparse set of model parameters. The first technique, the least absolute shrinkage and selection operator, uses the l(1) norm loss function to create a sparse linear model; the second, the support vector machine with radial basis kernel functions, uses the epsilon-insensitive loss function to create a sparse non-linear model. The techniques are applied to predict the response of the HIV-1 virus to 10 reverse transcriptase inhibitor and 7 protease inhibitor drugs. The genetic data are derived from the HIV coding sequences for the reverse transcriptase and protease enzymes. When tested by cross-validation with actual laboratory measurements, these models predict drug response phenotype more accurately than models previously discussed in the literature, and other canonical techniques described here. Key features of the methods that enable this performance are the tendency to generate simple models where many of the parameters are zero, and the convexity of the cost function, which assures that we can find model parameters to globally minimize the cost function for a particular training dataset. AVAILABILITY: Results, tables and figures are available at ftp://ftp.genesecurity.net. SUPPLEMENTARY INFORMATION: An Appendix to accompany this article is available at Bioinformatics online.

Amino Acid Sequence↗
PubMed1999

Statistical method to evaluate management strategies to decrease variability in operating room utilization: application of linear statistical modeling and Monte Carlo simulation to operating room management.

BACKGROUND: Operating room (OR) managers seeking to maximize labor productivity in their OR suite may attempt to reduce day-today variability in hours of OR time for which there are staff but for which there are no cases ("underutilized time"). The authors developed a method to analyze data from surgical services information systems to evaluate which management interventions can most effectively decrease variability in underutilized time. METHODS: The method uses seven summary statistics of daily workload in a surgical suite: daily allocated hours of OR time, estimated hours of elective cases, actual hours of elective cases, estimated hours of add-on cases, actual hours of add-on cases, hours of turnover time, and hours of underutilized time. Simultaneous linear statistical equations (a structural equation model) specify the relationship among these variables. Estimated coefficients are used in Monte Carlo simulations. RESULTS: The authors applied the analysis they developed to two OR suites: a tertiary care hospital's suite and an ambulatory surgery center. At both suites, the most effective strategy to decrease variability in underutilized OR time was to choose optimally the day on which to do each elective case so as to best fill the allocated hours. Eliminating all (1) errors in predicting how long elective or add-on cases would last, (2) variability in turnover or delays between cases, or (3) day-to-day variation in hours of add-on cases would have a small effect. CONCLUSIONS: This method can be used for decision support to determine how to decrease variability in underutilized OR time.

Humans↗
PubMed2004

Gamma regression improves Haseman-Elston and variance components linkage analysis for sib-pairs.

Existing standard methods of linkage analysis for quantitative phenotypes rest on the assumptions of either ordinary least squares (Haseman and Elston [1972] Behav. Genet. 2:3-19; Sham and Purcell [2001] Am. J. Hum. Genet. 68:1527-1532) or phenotypic normality (Almasy and Blangero [1998] Am. J. Hum. Genet. 68:1198-1199; Kruglyak and Lander [1995] Am. J. Hum. Genet. 57:439-454). The limitations of both these methods lie in the specification of the error distribution in the respective regression analyses. In ordinary least squares regression, the residual distribution is misspecified as being independent of the mean level. Using variance components and assuming phenotypic normality, the dependency on the mean level is correctly specified, but the remaining residual coefficient of variation is constrained a priori. Here it is shown that these limitations can be addressed (for a sample of unselected sib-pairs) using a generalized linear model based on the gamma distribution, which can be readily implemented in any standard statistical software package. The generalized linear model approach can emulate variance components when phenotypic multivariate normality is assumed (Almasy and Blangero [1998] Am. J. Hum Genet. 68: 1198-1211) and is therefore more powerful than ordinary least squares, but has the added advantage of being robust to deviations from multivariate normality and provides (often overlooked) model-fit diagnostics for linkage analysis.

Analysis of Variance↗
PubMed2005

Reduced folate carrier polymorphisms and neural tube defect risk.

The reduced folate carrier (RFCI) is essential for folate transport into cells. Low folate is an important cause of neural tube defects (NTDs), and a single-nucleotide polymorphism (H27R) (80G-->A) in the RFCI gene has been reported to be a NTD risk factor. We investigated H27R and a 61 bp tandem repeat polymorphism as potential risk factors for NTDs, using a large homogeneous Irish population by case/control comparison, log-linear analysis, and transmission disequilibrium testing. No association was found between NTDs and H27R in mothers [p = 0.23, odds ratio (OR) 0.87, 95% confidence interval (CI) 0.69-1.09], fathers (p = 0.11, OR 0.83, 95% CI 0.66-1.04), or cases (p = 0.36, OR 0.9, 95% CI 0.72-1.12) when compared to controls or through log-linear modeling for dominant or recessive effects or with the transmission disequilibrium test for preferential allele transmission. Using log-linear models, a significant protective case effect was seen for the 61 bp polymorphism (p = 0.0039, OR 0.21, 95% CI 0.05-0.85). When analyzed by genotype, individuals homozygous for a single copy of the 61 bp sequence were underrepresented in cases as compared to controls, although these results did not reach statistical significance (p = 0.081, OR 0.5, 95% CI 0.23-1.09, goodness of fit p = 0.42). We compared the frequencies of H27R and the 61 bp polymorphism in African-Americans and American-Caucasians. The frequencies of H27R polymorphism differed significantly between the two populations (p = 0.0001). This large study does not confirm previous reports that H27R is a risk factor for NTDs. The previously unstudied 61 bp tandem repeat, however, has a possible protective NTD effect in our Irish population. This requires confirmation in other studies.

Black or African American↗
PubMed2000

Diet high in monounsaturated fat does not have a different effect on arterial elasticity than a low-fat, high-carbohydrate diet.

OBJECTIVE: To compare the effects of a modified-fat diet high in monounsaturated fat, and a low-fat/high-carbohydrate diet on arterial elasticity. DESIGN: Randomized crossover design; each diet period was 1 month and a 2-week wash out period occurred in between. SUBJECTS/SETTING: Thirty healthy, free-living, nonsmoking men and women were recruited from the Melbourne, Australia, metropolitan region of Australia. Men were aged 35 to 55 years and postmenopausal women were aged 50 to 60 years and were not taking hormone replacement therapy. Twenty-eight subjects completed the study. INTERVENTION: Two diets of equal energy value: a modified-fat diet and a low-fat/high-carbohydrate diet; the modified-fat diet had 3 times more energy from monounsaturated fat. MAIN OUTCOME MEASURES: Arterial elasticity and serum lipoprotein concentrations. STATISTICAL ANALYSIS: The general linear model was used to investigate overall effect and any carryover or order effects. Paired t test and the general linear model were used to compare the results from the 2 diet periods. RESULTS: High-density lipoprotein cholesterol concentration was significantly higher on the modified-fat diet than on the low-fat/low-carbohydrate diet. Arterial elasticity and concentrations of total cholesterol, low-density lipoprotein cholesterol, and triglycerides were not significantly different on the 2 diets. APPLICATIONS/CONCLUSIONS: There is no evidence to favor a diet high in monounsaturated fat over a low-fat/high-carbohydrate diet because of an effect on arterial elasticity. Other changes in diet may be needed to cause a beneficial effect on arterial elasticity.

Adult↗
PubMed2002

Endoscopic internal thoracic artery dissection leads to significant reduction of pain after minimally invasive direct coronary artery bypass graft surgery.

BACKGROUND: The aim of this study was to evaluate postoperative pain levels after endoscopic versus conventional internal thoracic artery (ITA) dissection for minimally invasive direct coronary artery bypass graft surgery (MIDCABG) surgery. Results were compared with pain levels associated with conventional cardiac bypass operations through a median sternotomy. METHODS: Of 190 patients included in this prospective study, 24 patients had endoscopic ITA takedown (MIDCABG-endo) using the da Vinci telemanipulator followed by a manual coronary anastomosis through a left minithoracotomy. A conventional MIDCABG operation (MIDCABG-conv) was performed in 73 patients with ITA preparation under direct vision. Postoperative pain levels after conventional CABG through a median sternotomy (CABG-conv, n = 93) served as controls. A standarized questionnaire including visual analog scale (VAS) was used for prospective pain assessment from POD 1 to 7. RESULTS: Pain levels (VAS) declined in all groups from POD 1 to 7. Overall pain levels were significantly lower in the MIDCABG-endo group as compared with MIDCABG-conv and CABG-conv groups, respectively (p < 0.001, general linear model). There was no significant difference between the MIDCABG-conv and CABG-conv (p = not significant, general linear model) groups. Furthermore, patients after MIDCABG-endo required fewer nonsteroidal anti-inflammatory drugs and opioid medications, postoperatively. CONCLUSIONS: An endoscopic ITA takedown in MIDCABG surgery leads to significantly reduced postoperative pain levels possibly because of less rib retraction.

Aged↗
PubMed2003

Environmental correlates to behavioral health outcomes in Alzheimer's special care units.

PURPOSE: We systematically measured the associations between environmental design features of nursing home special care units and the incidence of aggression, agitation, social withdrawal, depression, and psychotic problems among persons living there who have Alzheimer's disease or a related disorder. DESIGN AND METHODS: We developed and tested a model of critical health-related environmental design features in settings for people with Alzheimer's disease. We used hierarchical linear modeling statistical techniques to assess associations between seven environmental design features and behavioral health measures for 427 residents in 15 special care units. Behavioral health measures included the Cohen-Mansfield physical agitation, verbal agitation, and aggressive behavior scales, the Multidimensional Observation Scale for Elderly Subjects depression and social withdrawal scales, and BEHAVE-AD (psychotic symptom list) misidentification and paranoid delusions scales. Statistical controls were included for the influence of, among others, cognitive status, need for assistance with activities of daily living, prescription drug use, amount of Alzheimer's staff training, and staff-to-resident ratio. Although hierarchical linear modeling minimizes the risk of Type II-false positive-error, this exploratory study also pays special attention to avoiding Type I error-the failure to recognize possible relationships between behavioral health characteristics and independent variables. RESULTS: We found associations between each behavioral health measure and particular environmental design features, as well as between behavioral health measures and both resident and nonenvironmental facility variables. IMPLICATIONS: This research demonstrates the potential that environment has for contributing to the improvement of Alzheimer's symptoms. A balanced combination of pharmacologic, behavioral, and environmental approaches is likely to be most effective in improving the health, behavior, and quality of life of people with Alzheimer's disease.

Aged↗
PubMed2004

Instant of chair-rise lift-off can be predicted by foot-floor reaction forces.

OBJECTIVE: To develop a predictive model of the lift-off event during chair rise in healthy subjects, using foot-floor reaction forces. BACKGROUND: An important event during chair rise is lift-off from the seat: the transition from the inherently stable three-point contact to the unstable two-point contact. There is no consistent or generally agreed upon method for estimating the time of lift-off when an instrumented seat is unavailable. METHODS: Twenty healthy volunteers were divided into a testing set and training set. Each subject performed repeated chair rise trials at different speeds. Seat-floor and foot-floor forces, recorded with two force platforms, were used to develop a model of the lift-off event. RESULTS: The magnitude of the vertical foot-floor reaction at lift-off (F0VF) was linearly related (R2 = 0.71, P < 0.001) to the peak vertical foot-floor reaction force (FMVF). A linear model was developed for the training group, which enabled prediction of lift-off time for the testing group with an absolute average error of 6 ms (about 1 data frame at 150 Hz). The linear model derived for the entire sample was: F0VF = 28.14 + FMVF * (0.6434). CONCLUSIONS: The lift-off event for healthy subjects performing chair rise can be accurately predicted from foot-floor reaction forces, without requiring an instrumented seat.

Body Weight↗
PubMed2002

Evaluation of command algorithms for control of upper-extremity neural prostheses.

Five new command control algorithms were created to enable increased control over grasp force in upper-extremity neural prostheses. Most of these algorithms took advantage of the ability to lock or assign a steady command value to the hand neural prosthesis. Five able-bodied subjects tested the algorithms by using a shoulder controller that controlled a video-simulated hand to repeatedly complete a consistent evaluation task. A generalized estimating equations-based linear model was used to analyze the data. The algorithms were ranked via contrast analyses between the coefficient values from the linear model of the proportional control with lock algorithm, which is the algorithm presently used in neural prostheses, and each of the other algorithms. The algorithms that allowed adjustment of the command value after the hand was locked as well as algorithms that allowed a decrease in controller gain after the hand was locked performed better than the proportional control with lock algorithm. Algorithms that changed command as a function of time performed worse than the proportional control with lock algorithm. Further, the computer-based video simulator proved to be useful as a first-pass evaluation tool for neural prosthesis control.

Adult↗
PubMed1988

The statistical analysis of treatment effects in 24-hour ambulatory blood pressure recordings.

This paper presents a statistical analysis of treatment effects in 24-hour ambulatory blood pressure recordings. The statistical models account for circadian rhythms, subject effects, and the effects of treatment with drugs or relaxation therapy. In view of the heterogeneity of the subjects, we fit a separate linear model to the data of each subject, use robust statistical procedures to estimate the parameters of the linear models, and trim the data on a subject by subject basis. We use a meta-analytical method to combine the results of all subjects in the study.

Activity Cycles↗
PubMed1986

Quantification of the initial phases of rapid brain enlargement in the chick embryo.

Rapid brain enlargement requires a hydraulic mechanism in the chick embryo. Such a mechanism involves a closed, fluid-filled system that generates positive pressure. For the chick embryo this study determined when rapid brain enlargement begins, assessed the relative contributions of cavity expansion and tissue growth to overall brain enlargement, and evaluated mathematical models of overall brain enlargement and expansion and growth of the component parts. Three to five embryos were collected at each Hamburger and Hamilton state (11, 12, 14, 16, and 18) and processed for paraffin serial sectioning. Brain growth was determined over a 24-hr period (stages 11-18) by calculating volumes from area measurements of sections of brains from individual embryos by using a computerized image-analysis system. Statistical analysis indicated that a linear model adequately described cavity expansion, and a linear model was rejected for the description of tissue growth and total brain enlargement. At the onset of brain enlargement, the cavity expands faster than the tissue grows; but after 12 hr the reverse is true. Initially (i.e., at stage 11), the cavity accounts for 60% of the total brain volume and tissue for 40%. At stages 12-16, cavity and tissue contribute 50% each. Finally at stage 18, cavity accounts for 55% and tissue for 45%. In order to better distinguish changes in cavity expansion and tissue growth over the 24-hr period studied, this period was divided into four intervals (I-IV). The rates of both cavity expansion and tissue growth increase between intervals I and II, decrease between intervals II and III, and increase between intervals III and IV.(ABSTRACT TRUNCATED AT 250 WORDS)

Animals↗
PubMed1997

Genetic analysis of discrete reproductive traits in sheep using linear and nonlinear models: I. Estimation of genetic parameters.

Repeated records on fertility, litter size, and ovulation rate of Rambouillet ewes and on fertility and litter size of Finnsheep ewes were used to estimate heritabilities and repeatabilities with linear and nonlinear sire and animal models. Linear sire (LSM) and animal models were used with all traits. Nonlinear models were the threshold, Poisson, and negative binomial. Threshold sire (TSM) and animal models were used with all traits. Litter size and ovulation rate were analyzed also with Poisson sire and animal models and with negative binomial sire and animal models. Variance components for linear models were estimated using REML; in the threshold, Poisson, and negative binomial, they were estimated using approximate marginal maximum likelihood. Poisson and negative binomial analyses yielded results difficult to interpret due to problems in variance component estimation. Animal models resulted in slightly greater estimates of heritability for fertility than did sire models, but ovulation rate heritability estimates from sire models were much greater than estimates form animal models. Differences between sire and animal models for heritability estimates for litter size were not consistent. Threshold models resulted in higher heritability estimates for all traits in both breeds and with both sire and animal models. In general, repeatabilities were consistent across models. For example, LSM (TSM) repeatabilities were .10 (.14) for fertility, .20 (.25) for litter size, and .25 (.29) for ovulation rate in the Rambouillet, and .17 (.17) for fertility and .11 (.13 for litter size in the Finnsheep.

Aging↗
PubMed2002

Development and validation of a non-linear IVIVC model for a diltiazem extended release formulation.

To develop and validate internally an in vitro-in vivo correlation (IVIVC) for a diltiazem multi-particulate bead extended release formulation. In vitro dissolution of diltiazem capsules was examined using the following methods: USP Apparatus II (paddle) at 100 rpm and USP Apparatus III at 30 dpm. Seven healthy subjects received three diltiazem formulations (90 mg): slow (S), moderate (M), fast (F) releasing and an oral solution (90 mg). Serial blood samples were collected over 48 h and analyzed by a validated HPLC assay using ultraviolet detection. The f(2) metric (similarity factor) was used to analyze the dissolution data. Linear and non-linear (quadratic, cubic, and sigmoid functions) correlation models were developed using pooled fraction dissolved (FRD) and fraction absorbed (FRA) data from various combinations of the formulations. Predicted diltiazem concentrations were obtained by convolution of the in vivo dissolution rates. Prediction errors were estimated for C(max) and AUC to determine the validity of the correlation. Apparatus II using purified water was found to be the most discriminating dissolution method. Significant intersubject (CV%>50) was observed for C(max) and AUC. The quadratic M/F IVIVC model provided a significant relationship between FRD and FRA when using either two or three of the formulations. An average percent prediction error for C(max) and AUC for all formulations was 12.4% and 9.2%, respectively. The prediction errors observed for C(max) and AUC suggest that the predictability of the quadratic IVIVC model is inconclusive, as such, external validation studies are required.

Calcium Channel Blockers↗
PubMed2006

A multivariate Bayesian model for assessing morbidity after coronary artery surgery.

INTRODUCTION: Although most risk-stratification scores are derived from preoperative patient variables, there are several intraoperative and postoperative variables that can influence prognosis. Higgins and colleagues previously evaluated the contribution of preoperative, intraoperative and postoperative predictors to the outcome. We developed a Bayes linear model to discriminate morbidity risk after coronary artery bypass grafting and compared it with three different score models: the Higgins' original scoring system, derived from the patient's status on admission to the intensive care unit (ICU), and two models designed and customized to our patient population. METHODS: We analyzed 88 operative risk factors; 1,090 consecutive adult patients who underwent coronary artery bypass grafting were studied. Training and testing data sets of 740 patients and 350 patients, respectively, were used. A stepwise approach enabled selection of an optimal subset of predictor variables. Model discrimination was assessed by receiver operating characteristic (ROC) curves, whereas calibration was measured using the Hosmer-Lemeshow goodness-of-fit test. RESULTS: A set of 12 preoperative, intraoperative and postoperative predictor variables was identified for the Bayes linear model. Bayes and locally customized score models fitted according to the Hosmer-Lemeshow test. However, the comparison between the areas under the ROC curve proved that the Bayes linear classifier had a significantly higher discrimination capacity than the score models. Calibration and discrimination were both much worse with Higgins' original scoring system. CONCLUSION: Most prediction rules use sequential numerical risk scoring to quantify prognosis and are an advanced form of audit. Score models are very attractive tools because their application in routine clinical practice is simple. If locally customized, they also predict patient morbidity in an acceptable manner. The Bayesian model seems to be a feasible alternative. It has better discrimination and can be tailored more easily to individual institutions.

Aged↗
PubMed2006

Correction of eddy-current distortions in diffusion tensor images using the known directions and strengths of diffusion gradients.

PURPOSE: To correct eddy-current artifacts in diffusion tensor (DT) images without the need to obtain auxiliary scans for the sole purpose of correction. MATERIALS AND METHODS: DT images are susceptible to distortions caused by eddy currents induced by large diffusion gradients. We propose a new postacquisition correction algorithm that does not require any auxiliary reference scans. It also avoids the problematic procedure of cross-correlating images with significantly different contrasts. A linear model is used to describe the dependence of distortion parameters (translation, scaling, and shear) on the diffusion gradients. The model is solved numerically to provide an individual correction for every diffusion-weighted (DW) image. RESULTS: The assumptions of the linear model were successfully verified in a series of experiments on a silicon oil phantom. The correction obtained for this phantom was compared with correction obtained by a previously published method. The algorithm was then shown to markedly reduce eddy-current distortions in DT images from human subjects. CONCLUSION: The proposed algorithm can accurately correct eddy-current artifacts in DT images. Its principal advantages are that only images with comparable signals and contrasts are cross-correlated, and no additional scans are required.

Algorithms↗