Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Linear Regression”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

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

At least 469 records · Page 26Linked to original sources
PubMedDate not supplied

Reference ranges of left ventricular cross-sectional echocardiographic measurements in adult men.

Reference ranges for left ventricle cross-sectional echocardiographic measurements were determined in 67 healthy Dutch men. Age, weight, height, and heart rate were used as the determinants. The images were made with the apical long-axis view, and the calculations were done with the area-length method. The end-diastolic volume, end-systolic volume, stroke volume, and ejection fraction were measured for six consecutive cardiac cycles in every subject and were averaged. Data were analyzed by means of both simple linear regression and multiple linear regression after logarithmic transformation of all measurements. Weight proved to be the best predictor of the echocardiographic end-diastolic volume, end-systolic volume, and stroke volume. The left ventricular ejection fraction could be regarded as independent of the determinants studied. Nomograms of the 5th, 50th and 95th percentile limits were made of the echocardiographic parameters versus weight.

Adult↗
PubMed2002

Preload and haemodynamic assessment during liver transplantation: a comparison between the pulmonary artery catheter and transpulmonary indicator dilution techniques.

BACKGROUND AND OBJECTIVE: Liver transplantation is characterized by several changes in intravascular blood volume due to vasodilatation based on neurohumoral mediators, intraoperative bleeding and anaesthesia technique effects. Today, with the transpulmonary indicator dilution technique, cardiac index (CIart) can be evaluated and preload assessed in terms of the intrathoracic blood volume index (ITBVI). The aim was to analyse in patients undergoing liver transplantation two preload variables, pulmonary artery occlusion pressure (PAOP) and ITBVI with respect to cardiac index (CIpa) and stroke volume index (SVIpa), the correlation between ITBVI and PAOP, and secondary the relationship between the changes (delta) of ITBVI and PAOP and the changes of CIpa and SVIpa, and the relationships between deltaITBVI and deltaPAOP. The reproducibility and precision of all CIart and CIpa measurements were also evaluated. METHODS: A prospective study was performed in 60 patients monitored with a pulmonary artery catheter and with the PiCCO system. The variables were evaluated with a linear regression model. RESULTS: Linear regression analysis between ITBVI-CIpa and ITBVI-SVIpa were r2 = 0.47 (P < 0.0001) and r2 = 0.55 (P < 0.0001) respectively, while PAOP poorly correlated to CIpa (r2 = 0.02), SVIpa (r2 = 0.015) and ITBVI (r2 = 0.002). Only changes in ITBVI were correlated with changes in CIpa (delta1, r2 = 0.37; delta2, r2 = 0.32), and SVIpa (delta1, r2 = 0.60; delta2, r2 = 0.47). The mean bias between CIart and CIpa was 0.13 L min(-1) m(-2) (2 SD = 1.04L min(-1) m(-2)) (r2 = 0.86, P < 0.0001). CONCLUSIONS: In comparison with PAOP, ITBVI seems a more reliable indicator of cardiac preload in patients undergoing liver transplantation.

Adult↗
PubMed1995

Lifetime leisure exercise and osteoporosis. The Rancho Bernardo study.

Between 1988 and 1991, the relation between leisure time physical activity, bone mineral density (BMD), and osteoporotic fracture was evaluated in a cohort of community-dwelling California adults (1,014 women and 689 men) with a mean age of 73 years. By means of a modified Paffenbarger questionnaire, participants were asked to report exercise from the past year and to recall their level of exercise during three other periods: the teenage years, age 30 years, and age 50 years. The survey asked the number of times strenuous (e.g., jogging), moderate (e.g., fast walking), or mild (e.g., golfing) exercise was undertaken in an average week. A summary score was constructed to represent lifetime exercise. Analysis of the exercise-fracture and exercise-BMD associations were performed using logistic and linear regression analyses, respectively. Linear regression models were controlled for age, body mass index, sex, diagnosis of arthritis, dietary calcium intake, and use of cigarettes, alcohol, thiazides, and estrogen (women only). No association between current or former exercise and BMD at the radius, wrist, or spine was found. A positive association between current exercise and BMD was found at the total hip (p = 0.001) and at each hip component--greater trochanter (p = 0.02), intertrochanter (p = 0.001), and femoral neck (p = 0.02). Mean hip bone densities of strenuous (p = 0.004) and moderate (p = 0.004) current exercisers were higher than those of mild or less than mild exercisers. Lifetime exercise was also positively associated with BMD of the total hip (p = 0.008) and hip components, and demonstrated a borderline-significant association (p = 0.06) with spine BMD. At the hip, each pairwise comparison between the highest and lowest tertiles of lifetime exercise showed a significant difference (p < or = 0.007). Exercise was unassociated with minimal trauma fracture occurring at any site between 1972 and 1991. These data suggest a protective effect of current and lifelong exercise on hip BMD, but not on osteoporotic fracture, in older men and women.

Aged↗
PubMed1997

Outcome for preterm infants with germinal matrix hemorrhage and progressive hydrocephalus.

OBJECTIVE: An analysis of 76 preterm infants with Grade III or IV intracranial hemorrhage and surgically treated progressive hydrocephalus was undertaken to determine mortality, intellectual impairment, and motor deficit. METHODS: The variables examined were degree of prematurity, birth weight, sex, Apgar scores, extent of intracranial hemorrhage, seizures, age at time of initial placement of a ventricular catheter reservoir to control hydrocephalus, need to convert the reservoir to a ventriculoperitoneal shunt, timing of the conversion of the reservoir to a ventriculoperitoneal shunt, and number of shunt revisions. Outcome was assessed for statistical significance using hierarchical linear regression and logistic regression analyses. RESULTS: Linear regression analysis determined that mortality was best predicted, in order of importance, by extent of intracranial hemorrhage, number of shunt revisions, and birth weight (P < 0.0001, R = 0.79). Grade of hemorrhage, weight at birth, and presence of seizure activity were the most important determinants of motor outcome (P < 0.001, R = -0.78). CONCLUSIONS: Logistic regression analysis of the 41 long-term survivors determined that grade of hemorrhage was the most important variable in determining cognitive outcome (P < 0.0001), motor function (P < 0.0001), and presence of seizure activity (P < 0.001). A logistic model of survival determined that grade of hemorrhage and multiple shunt revisions (more than five) were the most important determinants (P < 0.0001) of survival. In conclusion, the overwhelming factor in determining outcome in this patient group was the extent of intracranial hemorrhage.

Apgar Score↗
PubMed1989

Development and evaluation of spectral classification algorithms for fluorescence guided laser angioplasty.

Laser angioplasty, or the ablation of atherosclerotic plaque using laser energy, has tremendous potential to expand the scope of nonsurgical treatment of obstructive vascular disease. Clinical laser angioplasty, however, has been hindered by an unacceptable risk of vessel perforation. Laser-induced fluorescence spectroscopy can discriminate atherosclerotic from normal artery and may therefore be capable of guiding selective plaque ablation. To assess the feasibility of utilizing spectral information to discriminate arterial tissue type, several classification algorithms were developed and evaluated. Arterial fluorescence spectra from 350 to 700 nm were obtained from 100 human aortic specimens. Seven spectral classification algorithms were developed with the following techniques: multivariate linear regression, stepwise multivariate linear regression, principal components analysis, decision plane analysis, Bayes decision theory, principal peak ratio, and spectral width. The classification ability of each algorithm was evaluated by its application to the training set and to a validation set containing 82 additional spectra. All seven spectral classification algorithms prospectively classified atherosclerotic and normal aorta with an accuracy greater than 80 percent (range: 82-96 percent). Laser angioplasty systems incorporating spectral classification algorithms may therefore be capable of detection and selective ablation of atherosclerotic plaque.

Algorithms↗
PubMed2006

Low testosterone levels are associated with carotid atherosclerosis in men.

OBJECTIVE: To study the relationship between endogenous sex hormone levels and intima-media thickness (IMT) of the carotid artery measured by ultrasonography. DESIGN: Population-based cross-sectional study. METHODS: Sex hormone levels measured by immunoassay, anthropometric measurements and IMT was studied in 1482 men aged 25-84 years participating in the 1994-1995 Tromsø study. The data were analysed with partial correlation, multiple linear regression and logistic regression analysis. RESULTS: Linear regression models showed that total testosterone and sex hormone-binding globulin levels, but not calculated free testosterone, serum oestradiol or dehydroepiandrosterone sulphate levels were inversely associated with the age-adjusted IMT (P = 0.008 and P < 0.001 respectively). These associations were independent of smoking, physical activity, blood pressure and lipid levels, but were not independent of body mass index (BMI). Excluding men with cardiovascular disease (CVD) did not materially change these results. In a logistic regression model adjusted for the confounding effect of CVD risk factors, men with testosterone levels in the lowest quintile (<9.0 nmol L(-1)) had an independent OR = 1.51 (P = 0.015) of being in the highest IMT quintile. CONCLUSIONS: We found an inverse association between total testosterone levels and IMT of the carotid artery in men that was present also after excluding men with CVD, but was not independent of BMI. The clinical relevance of this, however, is uncertain and needs to be investigated in a clinical setting.

Adult↗
PubMed1995

DNA index and S phase fraction in uveal malignant melanomas.

AIMS: To predict 5 year survival in patients with uveal malignant melanomas DNA indices were studied. METHODS: Using 45 paraffin embedded uveal malignant melanomas, the DNA index and S phase fraction of each tumour were the predictor variables recorded. RESULTS: Using the Cox proportional hazards model, aneuploid tumours and tumours which had an S phase fraction greater than 4% were significant predictors of early death. In order to demonstrate a biological gradient between a larger DNA index and shorter survival time, linear regression and transformed linear regression models were used. However, no such gradient could be demonstrated. CONCLUSION: Although this study shows promise for the use of DNA studies in the prognosis of uveal malignant melanoma, the exact role of these techniques remains to be determined.

Adult↗
PubMed2001

Effects of tallow on the energy metabolism of wethers fed barley finishing diets.

A balance trial was conducted to titrate the effects of tallow on the energy metabolism of wethers fed barley finishing diets. Six dietary levels of tallow (0, 2, 4, 6, 8, or 10%) in a barley finishing diet were fed to six crossbred wethers (35+/-1.1 kg) in a randomized complete block design. Diets were 73% barley, 10% tallow and(or) bentonite, 10% alfalfa pellets, and 7% supplement. There was no effect of tallow level on OM intake (1,103.1+/-51 g/d), OM digestibility (84+/-0.9%), GE digestibility (83+/-1.1%), or cell solubles digestibility (84.2+/-1.2%). The level of tallow quadratically decreased ADF digestibility (P < 0.05), methane emissions, and methane energy as a percentage of GE P < 0.01). There were linear increases in dietary GE (megacalories per kilogram of OM [P < 0.01]), dietary DE (megacalories per kilogram of OM [P < 0.05]), and dietary ME (megacalories per kilogram of OM [P < 0.01]), as dietary tallow increased. Numbers of ruminal protozoa (Entodinium spp. and Polyplastron sp.) decreased linearly (P < 0.05) with increased level of tallow. The energy value of tallow (calculated by difference) was low. The total-tract fatty acid digestibility of tallow was calculated by linear regression, without intercept, after accounting for the fatty acids digested from the base diet (0% tallow fed to a wether in a period). Fatty acids of the same carbon length were pooled for the regression analysis. All linear regressions were significant (P < 0.10) indicating no effect of tallow level on fatty acid digestibility. Lauric acid had low digestibility. The high digestibility of all C16 (89%) and C18 (104%) fatty acids suggests an effect of tallow on endogenous and microbial fatty acid excretion. Fatty acid digestibility was probably a minor contributor to the low energy content of tallow, calculated by difference, in these diets.

Animal Feed↗
PubMed2004

Methamphetamine abuse during pregnancy and its health impact on neonates born at Siriraj Hospital, Bangkok, Thailand.

To ascertain the impact of intrauterine methamphetamine exposure on the overall health of newborn infants at Siriraj Hospital, Bangkok, Thailand, birth records of somatic growth parameters and neonatal withdrawal symptoms of 47 infants born to methamphetamine-abusing women during January 2001 to December 2001 were compared to 49 newborns whose mothers did not use methamphetamines during pregnancy. The data on somatic growth was analyzed using linear regression and multiple linear regression. The association between methamphetamine use and withdrawal symptoms was analyzed using the chi-square. Home visitation and maternal interview records were reviewed in order to assess for child-rearing attitude, and psychosocial parameters. Infants of methamphetamine-abusing mothers were found to have a significantly smaller gestational age-adjusted head circumference (regression coefficient = -1.458, p < 0.001) and birth weight (regression coefficient = -217.9, p < or = 0.001) measurements. Methamphetamine exposure was also associated with symptoms of agitation (5/47), vomiting (11/47) and tachypnea (12/47) when compared to the non-exposed group (p < 0r =0.001). Maternal interviews were conducted in 23 cases and showed that: 96% of the cases had inadequate prenatal care (<5 visits), 48% had at least one parent involved in prostitution, 39% of the mothers were unwilling to take their children home, and government or non-government support were provided in only 30% of the cases. In-utero methamphetamine exposure has been shown to adversely effect somatic growth of newborns and cause a variety of withdrawal-like symptoms. These infants are also psychosocially disadvantaged and are at greater risk for abuse and neglect.

Adolescent↗
PubMedDate not supplied

Methodological factors in the variability of lung volume and specific airway resistance measured by body plethysmography.

Thoracic gas volume (TGV) and specific airway resistance (sRaw) are commonly measured using pressure type and flow type body plethysmographs. Within-subject variability of the data, defined as the coefficient of variation of eight to ten measurements during the same session, was assessed with the two kinds of instruments and compared in fifteen normal subjects. The reproducibility of data obtained several days apart was also compared. All measurements were made in a 480-l body chamber, which could be used in both the pressure and the flow mode. The signals were processed digitally using three different algorithms: 1) simple linear regression (LR); 2) linear regression with drift correction achieved by adding to, or subtracting from the plethysmographic signal a term proportional to time (LRC); 3) Fourier analysis (FFT). Within-subject variability of TGV was much larger with flow than with pressure plethysmography when the signals were processed by LR (14.5 +/- 7.5 vs 6.3 +/- 3.0%; p less than 0.001), but almost the same using LRC (6.7 +/- 3.2 vs 5.4 +/- 2.7%) and FFT (6.1 +/- 2.4 vs 5.0 +/- 2.4%). For sRaw, variabilities were larger and less influenced by methodological factors. Adequate digital processing may therefore largely remedy the inherently greater variability of TGV measurements with flow plethysmographs.

Adult↗
PubMed2003

Calculated follicle deviation using segmented regression for modeling diameter differences in cattle.

Segmented linear regression alone or in combination with simple linear regression was evaluated as an objective method to calculate the beginning of follicle deviation by modeling the sequential (Experiment 1) and non-sequential or single-point (Experiment 2) differences in diameter between the future dominant (F1) and largest subordinate (F2) follicles of Wave 1 in cattle. The segmented regression consisted of Segment 1 representing the common growth phase, Segment 2 representing the period of dominance, and a Join Point connecting the two segments and representing the end of the common growth phase and the beginning of deviation. The model was fit to the diameter differences for each heifer in Experiment 1 (n=15) and the group of heifers in Experiment 2 (n=40). The optimal Join Point value that corresponded to the maximum R(2) was designated the calculated hour (Experiment 1) or diameter of F1 (Experiment 2) at the beginning of deviation. In Experiment 1, simple linear regression was used to calculate the corresponding diameter of F1 at the beginning of deviation. Observed deviation was determined by inspection of the diameter profiles of F1 and F2 for comparison to calculated deviation. In Experiment 1, the observed method determined the beginning of deviation in 80% of the heifers, whereas, the regression method calculated deviation in 93% of the heifers including two of the three heifers in which observed deviation was not discernable (no significant difference between methods). The mean hours of deviation after wave emergence (Hour 0) and diameters of F1 at the corresponding hours were not significantly different between the observed (62 h and 8.4 mm) and calculated (61 h and 8.8 mm) methods. In Experiment 2, the diameter of F1 at the beginning of calculated deviation was 8.2 mm. The results indicated that the segmented regression model can provide an objective and more accurate alternative to estimate follicle deviation, especially when observed deviation is obscured by the complexity of follicle development in some waves.

Animals↗
PubMed1991

[Determinants of mitral inflow velocity profiles in relation to left ventricular volume change: evaluations by pulsed Doppler echocardiography and left ventriculography].

We evaluated relationships between pulsed Doppler echocardiographic (PDE) parameters of flow velocity profiles across the mitral orifice and left ventriculographic (LVG) parameters of left ventricular volume changes. Subjects consisted of 19 patients with coronary artery disease and 12 patients with chest pain syndrome. Peak flow velocities at the rapid filling (E) and atrial contraction (A) were measured by PDE. Time constant of left ventricular relaxation (T), left ventricular minimum pressure (LVPmin), LV end-diastolic pressure (LVEDP) and pulmonary capillary wedge V wave pressure (PCW-V) were measured during cardiac catheterization. From the analysis of LVG, the rapid filling fraction (RFF), and the atrial filling fraction (AFF) were obtained. The left ventricular chamber stiffness (K) was identified by the analysis of the pressure-volume relationship of the left ventricle. We investigated the relationship between A/E and AFF/RFF by univariate linear regression analysis. We then performed stepwise multivariate linear regression analysis to predict E, A, A/E, RFF, AFF and AFF/RFF by the variables of left ventricular filling, i.e., T, LVPmin, LVEDP, PCW-V, K, heart rate (at the examination of PDE or LVG), mean arterial blood pressure (at PDE or LVG) and age. The A/E correlated significantly with AFF/RFF (r = 0.50, p < 0.01). The results of the multivariate linear regression analyses were as follows: [sequence: see text] The correlation of A/E and AFF/RFF were explained by some variables, except the variable T. The results of uni- and multivariate linear regression analyses showed that factors affecting the flow velocity profile across the mitral orifice did not account for the left ventricular volume changes. We also observed that, even in subjects with coronary heart disease, aging is a main factor that influences peak flow velocity at atrial contraction (A).

Adult↗
PubMed2005

A new approach to correct the QT interval for changes in heart rate using a nonparametric regression model in beagle dogs.

INTRODUCTION: Over- and/or under-correction of QT intervals for changes in heart rate may lead to misleading conclusions and/or masking the potential of a drug to prolong the QT interval. This study examines a nonparametric regression model (Loess Smoother) to adjust the QT interval for differences in heart rate, with an improved fitness over a wide range of heart rates. METHODS: 240 sets of (QT, RR) observations collected from each of 8 conscious and non-treated beagle dogs were used as the materials for investigation. The fitness of the nonparametric regression model to the QT-RR relationship was compared with four models (individual linear regression, common linear regression, and Bazett's and Fridericia's correlation models) with reference to Akaike's Information Criterion (AIC). Residuals were visually assessed. RESULTS: The bias-corrected AIC of the nonparametric regression model was the best of the models examined in this study. Although the parametric models did not fit, the nonparametric regression model improved the fitting at both fast and slow heart rates. DISCUSSION: The nonparametric regression model is the more flexible method compared with the parametric method. The mathematical fit for linear regression models was unsatisfactory at both fast and slow heart rates, while the nonparametric regression model showed significant improvement at all heart rates in beagle dogs.

Administration, Oral↗
PubMed1981

Postnatal endocardial fibroelastosis of the valve of the foramen ovale.

The valve of the foramen ovale of the heart changes at birth from a prenatal shunting device for the transseptal flow of systemic venous return to a postnatal active barrier to backflow of left atrial blood at higher pressure. To investigate the morphologic response to this functional change in the valve of the foramen ovale, we examined 200 hearts from embryos, fetuses and autopsy patients spanning over a millionfold range of heart weights (0.001 to 1,230 g). The relation of valve thickness to log heart weight was studied by a stepwise linear regression procedure that fitted the data to two linear regressions, one including observations made before birth and the other including observations made after birth. Analysis of variance tables demonstrated that the data are significantly better explained (p < 0.001) by two linear regressions that abruptly separate from one another at the time of birth than by a single linear regression. Histologic study showed that the prenatal valve consists of a thin myocardial layer with delicate endocardium. The postnatal thickening of the valve tissue occurs mainly by fibroelastotic thickening of the endocardium. The endocardial fibroelastosis of the valve develops after it closes and becomes a component of the left atrial wall. The observations support the concept that fibroelastosis is a nonspecific response of the lining of the cardiovascular system to increased mural tension.

Adolescent↗
PubMed1981

[Regression analysis of the heart rate increase in children during linearly increasing work load on a bicycle ergometer].

E. A. Müller (1950) has introduced a procedure for the measurement of the "Leistungspulsindex" (LPI) based on the determination of a linear regression between heart rate increase and work load. For this purpose the work load on a bicycle ergometer is increased in proportion to time (about 10 W/min) and the heart rate is recorded continuously. We have performed measurements in children (10--11 years of age) and found that the above mentioned regression is markedly nonlinear and can be described best by two overlapping linear regression lines with different slopes (b11 and b12). We propose that under these conditions the regression coefficient b11 of the first part of the relation can be assumed as LPI. In our analysis we found the linear part of the heart rate to increase between the end of min 2 and min 6 (+/- 1.4) of the experiment in boys. In girls we found the linear part of the regression only until min 3.8 (+/- 1.4). One consequence of our results is related to the calculation of the PWC170 by linear extrapolation which appears to be impossible in all cases in which the relation between work load and heart rate is nonlinear. It seems possible to relate that part of the heart rate increase where the slope suddenly changes to the endurance limit.

Child↗
PubMed2004

Coronary artery calcification and cardiovascular risk factors: impact of the analytic approach.

Coronary artery calcification (CAC) may help identify novel risk factors for coronary atherosclerosis. However, analysis of CAC is challenging because of the distribution of CAC in the population. This has resulted in difficulty in interpreting and comparing results across studies. We applied several analytic approaches to CAC data in order to determine the impact of analytic methods on the association with established cardiovascular risk factors in 914 asymptomatic subjects in the Study of Inherited Risk Factors for Coronary Atherosclerosis. Multivariable analyses included: (1) linear regression of different transformations of CAC scores; (2) tobit regression of the log of (CAC + 1); (3) logistic regression using CAC zero as a cut-point; and (4) ordinal logistic regression using CAC categories. Linear regression of the log CAC scores and logistic regression of CAC zero cut-point failed to detect associations with some risk factors. In contrast, linear and tobit regression of the log (CAC + 1) and ordinal regression of CAC categories identified more associations and provided consistent results. Commonly applied methods of CAC analysis may fail to detect associations with cardiovascular risk factors. We present analytic approaches that are likely to provide consistent results and recommend the use of at least two distinct multivariable methods.

Adult↗
PubMed1999

Measuring breathlessness during histamine challenge: a simple standardized procedure in asthmatic patients.

One of the problems in research on symptom perception during histamine challenge has been the difficulty in finding both a valid and practical parameter of the "perceptiveness" for bronchoconstriction in a subject. The purpose of this study was to validate whether the slope in the linear regression model between stimulus and sensation during histamine challenge is an appropriate index for the "perceptiveness" for bronchoconstriction by comparing it with the classical Stevens' law. One hundred and thirty-four asthmatic patients were included in the study and underwent a bronchial challenge with histamine. The relationship between the change in visual analogue scale (VAS) values and the change in forced expiratory volume in one second (FEV1) as percentage of baseline value was analysed by determining both the exponent n in deltaVAS=k x (%deltaFEV1)n and the slope alpha in deltaVAS=k + alpha(%deltaFEV1). The best-fitting line of both the exponential and the linear regression model were determined by the least-squares method in which the percentage explained variation (R2) was compared. The median value of R2 of the exponential regression line and the linear regression line was 0.76 and 0.83, respectively, and significantly different. The Spearman rank correlation coefficient between exponent n in the exponential model deltaVAS=k x (%deltaFEV1)n and the slope alpha in the linear regression model deltaVAS=k + alpha(%deltaFEV1) was 0.87 (95% confidence interval 0.83-0.91). On the basis of the results, it was concluded that the linear regression coefficient can be used as a valid expression to describe the "perceptiveness" of an asthmatic subject instead of Stevens' power function during histamine challenge.

Adult↗
PubMed2003

Non-linear heteroscedastic regression model for determination of methotrexate in human plasma by high-performance liquid chromatography.

Generalized least squares regression with variance function estimation was used to derive the calibration function for measurement of methotrexate plasma concentration and its results were compared with weighted least squares regression by usual weight factors and also with that of ordinary least squares method. In the calibration curve range of 0.05 to 100 microM, both heteroscedasticity and non-linearity were present therefore ordinary least squares linear regression methods could result in large errors in the calculation of methotrexate concentration. Generalized least squares regression with variance function estimation worked better than both the weighted regression with the usual weight factors and ordinary least squares regression and gave better estimates for methotrexate concentration.

Calibration↗