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 1,063 records · Page 59Linked to original sources
PubMed1985

The effect of dextro-, levo-, and racemic verapamil on atrioventricular conduction in humans.

To study the dromotropic effects of dextro(D)- and levo(L)-verapamil on atrioventricular (AV) conduction in humans, we investigated the prolongation of the PR interval following intravenous administrations of each isomer and racemic preparation (D, 5, 25, and 50 mg; L, 5, 7.5, and 10 mg; racemic, 10 mg). The plasma drug concentration-effect relationship was analyzed by log-linear regression and the sigmoidal Emax model. The sigmoidal Emax model provided a significantly better fit for the data than log-linear regression (p less than 0.01). Maximum drug effect (Emax) and plasma drug concentration associated with 50% Emax (EC50) were calculated by means of the Emax model. The dromotropic potency of each isomer was assessed in terms of EC50 and the drug concentration associated with a 10% PR prolongation from the basal level calculated by the Emax model. The results demonstrated that L-verapamil was 10 and 18 times more potent than D-verapamil in terms of EC50 (D, 188.9 +/- 108.4 ng/ml; L, 17.7 +/- 11.3 ng/ml; p less than 0.05) and drug concentration associated with 10% PR prolongation (D, 166.6 +/- 48.1 ng/ml; L, 9.1 +/- 2.8 ng/ml; p less than 0.01), respectively. A stereospecific difference in plasma protein binding was observed (D, 93.7 +/- 2.2%; L, 88.5 +/- 1.6%; p less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)

Adult↗
PubMed2004

A long lifetime chemical sensor: study on fluorescence property of fluorescein isothiocyanate and preparation of pH chemical sensor.

The fluorescence property of fluorescein isothiocyanate (FITC) in acid-alkaline medium was studied by spectrofluorimetry. The characteristic of FITC response to hydrogen ion has been examined in acid-alkaline solution. A novel pH chemical sensor was prepared based on the relationship between the relative fluorescence intensity of FITC and pH. The measurement of relative fluorescence intensity was carried out at 362 nm with excitation at 250 nm. The excellent linear relationship was obtained between relative fluorescence intensity and pH in the range of pH 1-5. The linear regression equation of the calibration graph is F = 66.871 + 6.605 pH (F is relative fluorescence intensity), with a correlation coefficient of linear regression of 0.9995. Effects of temperature, concentration of FITC on the response to hydrogen ion had been examined. It was important that this chemical sensor was long lifetime, and the property of response to hydrogen ion was stable for at least 70 days. This pH sensor can be used for measuring pH value in water solution. The accuracy is 0.01 pH unit. The results obtained by the pH sensor agreed with those by the pH meter. Obviously, this pH sensor is potential for determining pH change real time in biological system.

Fluorescein-5-isothiocyanate↗
PubMed1995

Determining progressive visual field loss in serial Humphrey visual fields.

PURPOSE: To evaluate the three commercially available computerized statistical algorithms for determining progression of glaucomatous field loss in serial Humphrey fields. METHODS: Serial Humphrey fields of 102 ocular hypertensive and glaucomatous eyes were analyzed by linear regression analysis of Statpac, glaucoma change probability and Progressor Programme, and the results compared with a retrospectively determined clinical outcome. RESULTS: Linear regression analysis of Statpac identified progression in 11 field series, glaucoma change probability identified progression in 23 field series, and Progressor Programme identified progression in 27 field series. Clinical outcome identified much less-progression than any of the three algorithms, determining that only seven eyes showed deterioration. There was a wide variation of results from analysis of the same group of field series by the three different algorithms. In contrast to the other two algorithms, which detected progression in similar percentages of both ocular hypertension and glaucoma field series. Progressor Programme detected progression predominantly in glaucoma field series. CONCLUSIONS: There was a high degree of variability among the three different algorithms for determining visual field progression, and none of them correlated well with clinical impression.

Algorithms↗
PubMed1998

Blood pressure effects of long-term stimulant use in disorders of hypersomnolence.

Untoward cardiovascular effects have been implicated as a deterrent to long-term central nervous system (CNS) stimulant use in disorders of hypersomnolence. In this study, we reviewed the relationship between blood pressure and long-term stimulant use. Medical records of 54 patients with narcolepsy and idiopathic CNS hypersomnolence (ICH) were reviewed. The overall mean number of months of follow-up for the entire group was 45.6 (95% CI: 42-49). Both simple linear regression and multiple regression utilizing generalized estimating equations were used to show relationships between blood pressure (BP), time and other covariates. In the simple linear regression model, the average slope of the line of systolic BP (SBP) on time for the entire group was 0.06 (95% CI: -0.09, 0.13) and the line of diastolic BP (DBP) on time was 0.01 (95% CI: -0.05, 0.07). Two multiple regression equations were fitted for the continuous response variables SBP and DBP. Covariates in the model included: time, hypertension, weight at baseline, weight, SBP baseline (SBPBL), DBP baseline (DBPBL), high vs. low dose stimulant therapy and age at starting treatment. For SBP, the covariates weight at baseline, weight and SBPBL were significant (P < 0.05) predictors. For DBP, covariates reaching statistical significance (P < 0.05) included weight and DBPBL. There was no significant change in SBP or DBP over time in either model. Two different statistical models support the conclusion that there was no significant change in SBP or DBP over time in this population.

Adult↗
PubMed1997

Comparison of anti-factor Xa heparin activity and activated partial thromboplastin time in 2,773 plasma samples from unfractionated heparin-treated patients.

This study was conducted to evaluate two methods of determining the "therapeutic range" of the activated partial thromboplastin time (APTT) during unfractionated heparin treatment. Multiple fresh plasma samples from 694 patients treated with unfractionated heparin were tested simultaneously for APTT and anti-factor Xa heparin activity. The data were analyzed by linear regression and by a minimization-of-error technique to determine the more accurate APTT therapeutic range. The best-fit linear regression equation was obtained using logarithmically transformed APTT values. Using this equation, the heparin activity predicted by an APTT value had a 95% limit of agreement of +/- 0.39 U/mL compared with the actual heparin activity. The total clinical error rate of the APTT therapeutic range selected using the best-fit regression equation was 10%. An improvement to 8% was achieved using a method that chose the therapeutic range based on minimizing the APTT errors. Even 8% may be too high an error rate for best clinical results.

Evaluation Studies as Topic↗
PubMed1998

Valve Resistance in Mitral Stenosis: Its Determinants and its Role in the Evaluation of the Disease.

To evaluate the value and the determinants of valve resistance in mitral stenosis, 95 patients with pure mitral stenosis were examined by Doppler echocardiography during their clinical follow-up, measuring cavity dimensions, left ventricular function, mitral area (by planimetry and pressure half time), mean transmitral pressure gradient, aortic flow, and pulmonary artery systolic pressure. The mitral resistance was calculated as mean transmitral pressure gradient/aortic flow ratio. To graduate the severity of the morphological abnormalities in valvular structure, we used a point score system with evaluation of leaflet and subvalvular thickness, calcification, and valvular mobility. The functional class was determined according to NYHA classification. In this study, both mitral area (r = -0.79, P < 0.001 and r(p) = -0.60, P < 0.001) and mitral score (r = 0.68, P < 0.001 and r(p) = 0.25, P = 0.013) were independent determinants of mitral resistance. In multivariate analysis, mitral resistance and female gender were selected by multiple linear regression analysis as determinants of pulmonary artery systolic pressure, and mitral area and pulmonary artery systolic pressure were selected by logistic linear regression analysis as determinants of NYHA functional class. In patients with moderate or severe mitral stenosis, the estimated probability for III and IV NYHA functional class considering mitral area 1 cm(2) or below went from 51.1-86.4% when mitral resistance below or above 130 dynes.sec.cm(-5), respectively, was considered together. Thus, mitral valve resistance should be used as a complement to the mitral area method in assessment of mitral stenosis, adding the effects of the reduction in mitral area and the damage in mitral valve apparatus.

Journal Article↗
PubMed1979

Toxoplasma gondii-vertebrate cell interactions. II. The intracellular reproductive phase.

The reproduction of Toxoplasma gondii RH-strain in vertebrate cells was studied in a controlled-environment culture system. The lag period before reproduction and the doubling time of individual parasites were determined using a least-squares linear regression method of analysis which does not artificially constrain the data. In the majority of cases, the time intercept of the linear regression line was either zero, implying the lack of a lag phase before reproduction, or negative, implying the parasite had completed part of its reproductive cycle before entering the host cell. The mean doubling time of T. gondii is 10.9 h in bovine embryo skeletal muscle cells and 8.3 h in HeLa cells. This difference is not significant at the 5% level. The population doubling times of mouse-derived parasites is best described by a gamma distribution.

Animals↗
PubMed2006

Would artificial neural networks implemented in clinical wards help nephrologists in predicting epoetin responsiveness?

BACKGROUND: Due to its strong intra- and inter-individual variability, predicting the ideal erythropoietin dose is a difficult task. The aim of this study was to re-evaluate the impact of the main parameters known to influence the responsiveness to epoetin beta and to test the performance of artificial neural networks (ANNs) in predicting the dose required to reach the haemoglobin target and the monthly dose adjustments. METHODS: We did a secondary analysis of the survey on Anaemia Management in dialysis patients in Switzerland; a prospective, non-randomized observational study, enrolling 340 patients of 26 centres and in order to have additional information about erythropoietin responsiveness, we included a further 92 patients from the Renal Services of the Ente Ospedaliero Cantonale, Bellinzona, Switzerland. The performance of ANNs in predicting the epoetin dose was compared with that of linear regressions and of nephrologists in charge of the patients. RESULTS: For a specificity of 50%, the sensitivity of ANNs compared with linear regressions in predicting the erythropoietin dose to reach the haemoglobin target was 78 vs. 44% (P < 0.001). The ANN built to predict the monthly adaptations in erythropoietin dose, compared with the nephrologists' opinion, allowed to detect 48 vs. 25% (P < 0.05) of the patients treated with an insufficient dose with a specificity of 92 vs. 83% (P < 0.05). CONCLUSION: In predicting the erythropoietin dose required for an individual patient and the monthly dose adjustments ANNs are superior to nephrologists' opinion. Thus, ANN may be a useful and promising tool that could be implemented in clinical wards to help nephrologists in prescribing erythropoietin.

Adolescent↗
PubMed2003

Stroke volume variation as an indicator of fluid responsiveness using pulse contour analysis in mechanically ventilated patients.

Assessment of cardiac performance and adequate fluid replacement of a critically ill patient are important goals of a clinician. We designed this study to evaluate the ability of stroke volume variation (SVV), derived from pulse contour analysis, and frequently used preload variables (central venous pressure and pulmonary capillary wedge pressure) to predict the response of stroke volume index and cardiac index to volume replacement in normoventilated cardiac surgical patients. We studied 20 patients undergoing elective coronary artery bypass grafting. After the induction of anesthesia, hemodynamic measurements were performed before (T1) and subsequent to volume replacement by infusion of 6% hydroxyethyl starch 200/0.5 (7 mL/kg) with a rate of 1 mL x kg(-1) x min(-1). Except for heart rate, all hemodynamic variables changed significantly (P < 0.01) after volume loading. Linear regression analysis between SVV at baseline (T1) and DeltaSVV after volume application showed a significant correlation (r = -0.97; P < 0.01), whereas linear regression analysis between SVV (T1) and percentage changes of stroke volume index (r = 0.19) and cardiac index (r = 0.17) did not reveal a significant relationship between variables. The results of our study suggest that SVV derived from pulse contour analysis cannot serve as an indicator of fluid responsiveness in normoventilated cardiac surgical patients.

Adult↗
PubMed1986

Evaluation of elastase and alpha 1-proteinase inhibitor-elastase uptake by polymorphonuclear leukocytes and evidence of an elastase-specific receptor.

Neither resting nor stimulated isolated human polymorphonuclear leukocytes did bind or ingest preformed complexes of alpha 1-proteinase inhibitor and unlabeled/125I-labeled human leukocyte elastase. In contrast, granulocytes bound unlabeled/125I-labeled elastase and the extent of binding was reduced in the presence of respiratory burst stimulators, such as 4 beta-phorbol 12 beta-myristate 13 alpha-acetate, E. coli endotoxin, and N-formyl-L-methionyl-L-leucyl-L-phenylalanine. In association/dissociation and competition inhibition experiments it was demonstrated that granulocyte-elastase binding was specific and saturable. From Scatchard and non-linear regression analysis there was evidence of a two-class receptor model with independent binding sites. Calculated by the non-linear regression method assuming a two-class receptor model the characteristics of the high affinity/low capacity binding site were K1 = 216 +/- 129 X 10(6) l X mol-1 (means +/- s; n = 3) and R1 = 1.38 +/- 0.95 nmol X l-1 corresponding to 0.083 X 10(6) receptors per cell, whereas the low affinity/high capacity binding site had the characteristics K2 = 0.50 +/- 0.09 X 10(6) l X mol-1 and R2 = 237 +/- 103 nmol X l-1 corresponding to 14.3 +/- 6.2 X 10(6) receptors per cell.

Binding Sites↗
PubMed2000

Variation in proteolysis, sarcomere length, collagen content, and tenderness among major pork muscles.

The objectives of this experiment were to determine the extent of variation in proteolysis, sarcomere length, and collagen content among pork muscles and the association of those factors with tenderness variation among muscles at 1 d postmortem. Twenty-three white composite barrows were slaughtered and carcasses (66 kg) were chilled at 0 degrees C for 24 h. At 1 d postmortem, the longissimus lumborum, biceps femoris, semimembranosus, semitendinosus, and triceps brachii, long head were dissected from one side of each carcass and frozen. Trained sensory panelists evaluated tenderness, amount of connective tissue, juiciness, and pork flavor intensity of grilled (70 degrees C) chops on 8-point scales. Raw chops were used for total collagen content, sarcomere length, and the extent of desmin proteolysis. Tenderness ratings were highest (P < .05) for semitendinosus (7.2) and triceps brachii (7.1), followed by longissimus lumborum (6.4) and semimembranosus (5.7) and were lowest (P < .05) for biceps femorus (4.0). The simple correlations between longissimus lumborum tenderness and the tenderness of other muscles were .54 (semimembranosus), .34 (semitendinosus), .36 (triceps branchii), and .17 (biceps femorus). Total collagen was highest (P < .05) for biceps femorus (7.1 mg/g muscle), followed by triceps branchii (6.0 mg/g) and semitendinosus (5.3 mg/g), and lowest for semimembranosus (4.5 mg/g) and longissimus lumborum (4.1 mg/g). Sarcomere length was longest (P < .05) for semitendinosus (2.5 microm) and triceps branchii (2.4 microm), followed by semimembranosus (1.8 microm), longissimus lumborum (1.8 microm), and biceps femorus (1.7 microm). Proteolysis of desmin was greatest (P < .05) in longissimus lumborum (39.3%), followed by semimembranosus (21.0%) and biceps femoris (18.5%), then semitendinosus (.2%) and triceps brachii (.2%). Multiple linear regression using total collagen, sarcomere length, and proteolysis accounted for 57% of the variation in tenderness rating among all samples. Piecewise linear regression was used to account for the interaction of sarcomere length with proteolysis and collagen. This analysis accounted for 72% of the variation in tenderness rating. Variation in collagen, proteolysis, and sarcomere length and the degree of their interaction with one another determine the tenderness of individual muscles.

Animals↗
PubMed1995

Measurement of serum isopropanol and the acetone metabolite by proton nuclear magnetic resonance: application to pharmacokinetic evaluation in a simulated overdose model.

The purpose of this investigation was to 1) compare the performance of proton nuclear magnetic resonance spectroscopy to gas chromatography head-space analysis in the measurement of serum isopropanol and its metabolite, acetone, obtained during a simulated overdose, and 2) compare pharmacokinetic parameters obtained using the two analytical techniques. Three healthy volunteers ingested 0.6 mL/kg of 70% isopropanol and blood samples were obtained at baseline, 0.16, 0.33, 0.66, 1.0, 1.5, 2.0, 3.0, 4.0, 6.0, 8.0, 12.0, and 24.0 hours post-ingestion. Resulting sera were analyzed by gas chromatography head-space analysis and proton nuclear magnetic resonance spectroscopy for determination of isopropanol and acetone concentrations. A correlation between concentrations quantitated by gas chromatography head-space analysis versus proton nuclear magnetic resonance spectroscopy was determined using linear regression. Pharmacokinetic disposition parameters were determined from serum concentration-time data and compared using analysis of variance. For isopropanol, the linear regression equation which describes the relationship between gas chromatography head-space analysis and proton nuclear magnetic resonance spectroscopy was y = 1.041x - 2.180 (r2 = 0.995, p < 0.0001); for acetone, y = 1.022x - 0.946 (r2 = 0.984, p < 0.0001). Pharmacokinetic disposition parameters derived from the two analytical methods were comparable. Proton nuclear magnetic resonance spectroscopy can be used to rapidly quantitate serum isopropanol and acetone concentrations in the same sample when gas chromatography head-space analysis is unavailable. Also, proton nuclear magnetic resonance spectroscopy can be used to follow serial serum concentrations during an ingestion for the purpose of pharmacokinetic analysis.

1-Propanol↗
PubMed2001

Comparison of cardiac sympathetic nervous function with left ventricular function and perfusion in cardiomyopathies by (123)I-MIBG SPECT and (99m)Tc-tetrofosmin electrocardiographically gated SPECT.

UNLABELLED: The objective of this study was to clarify the relationship between cardiac sympathetic nervous function (CSNF) and left ventricular (LV) function and perfusion in hypertrophic cardiomyopathy (HCM) and dilated cardiomyopathy (DCM). METHODS: Thirty-eight cases (32 males, 6 females; mean age, 56 +/- 15 y), consisting of 5 healthy control subjects, 15 patients with DCM, and 18 patients with HCM, were studied with (123)I-metaiodobenzylguanidine (MIBG) and (99m)Tc-tetrofosmin SPECT. CSNF was evaluated from cardiac uptake and washout of MIBG, whereas LV perfusion and function were evaluated from tetrofosmin uptake and wall thickening on electrocardiographically gated SPECT. As quantitative parameters of global cardiac MIBG uptake and washout, the heart-to-mediastinum ratio (H/M) and percentage washout were calculated from early and delayed planar images. As quantitative regional parameters, the regional uptake and percentage washout of MIBG were calculated from SPECT images dividing the left ventricle into 12 segments. In the tetrofosmin study, the H/M and LV ejection fraction were calculated as the parameters of global LV perfusion and function. As quantitative regional parameters, the regional uptake and wall thickening were also calculated for the 12 myocardial segments using the quantitative gated SPECT software. Multiple linear regression analysis was performed to investigate the correlations between the parameters from the 2 studies. RESULTS: In DCM and HCM, multiple linear regression analysis of the regional parameters showed significant correlations between LV function and CSNF (P < 0.0001) and between LV perfusion and CSNF (P < 0.0001). According to the partial correlation coefficients, washout and early uptake of MIBG were the most significant factors for predicting LV function and LV perfusion, respectively. CONCLUSION: In cardiomyopathies, CSNF was closely related to LV function. The quantitative parameters of MIBG washout could reflect cardiac functional impairment. Early MIBG uptake might be determined by myocardial perfusion in cardiomyopathies.

3-Iodobenzylguanidine↗
PubMed2005

Utility of a standardized nursing terminology to evaluate dosage and tailoring of an HIV/AIDS adherence intervention.

PURPOSE: To illustrate the utility of a standardized nursing terminology to calculate the dosage of the Client Adherence Profiling-Intervention Tailoring (CAP-IT) and to determine the extent to which a tailored intervention was delivered to 117 persons with HIV/AIDS who participated in the experimental arm of a randomized controlled trial (RCT). METHODS: The intervention nurse assigned nursing diagnoses from the Home Health Care Classification (HHCC) based upon CAP scores. During the IT phase of CAP-IT, the nurse delivered and documented a tailored set of nursing interventions associated with the CAP and assigned nursing diagnoses. Hierarchical linear regression was used to evaluate the extent to which the number of interventions and intervention times were tailored to client needs. RESULTS: Linear regression models that included CAP scores and nursing diagnoses as predictor variables explained 53.2% of the variance in total number of interventions and 58.9% of the variance in intervention time. CONCLUSIONS: The use of the standardized nursing terminology enabled calculation of the intervention dose and documentation that a tailored intervention was delivered.

Adult↗
PubMed2004

[Clinical significance of stress reactivity for frequency of electric shock discharges in patients with implanted cardioverter-defibrillators].

BACKGROUND AND OBJECTIVE: Recent research indicates that discharges of implanted cardioverter-defibrillators (ICD) are triggered by psychosocial stress. This study examined the presence of patterns of psychophysiological stress reactivity in ICD-patients. Further more potential relationship between factors of stress reactivity and the frequency of implantable cardioverter-defibrillators discharges was investigated. PATIENTS AND METHODS: In 46 consecutive patients (38 male, 8 female; age: 26-80 years) with implantable cardioverter-defibrillators the following parameters of physiological reactivity were measured in a standardised experimental stress test: Heart rate, skin conductance, muscle tension and respiratory activity. In a standardised interview anxiety sensitivity, depression, desire for control, stress load, and attitude towards the defibrillator were evaluated and quantified. The relationship between the mentioned parameters were examined with factor analysis to reveal patterns of psychophysiological reactivity. Potential relationship between psychophysiological factors and frequency of shock delivery was verified with linear regression analysis. RESULTS: Factor analysis revealed four independent psychophysiological traits. The factors were named "Play Dead Reflex", "Heart Phobia", "Negativism" and "Irritability". Linear regression model showed significant correlation between "Play Dead Reflex" and discharge frequency. CONCLUSIONS: Results confirm the presence of specific psychophysiological stress reaction patterns in ICD-patients. The stress reaction pattern "Play Dead Reflex" could be discussed as a potential risk factor in developing life-threatening tachyarrhythmias leading to an increase in defibrillator discharges in patients with ICD. This hypothesis should be considered in prospective studies and psychosomatic treatment of concerned patients.

Adult↗
PubMedDate not supplied

Absolute TCD4+ counting by a minimalist dual-platform flow cytometric method.

UNLABELLED: The aim of this work was to compare the performance of an absolute TCD4+ counting method based on total WBC gating versus the standard lymphocyte (Ly) gating method, in order to develop a flow cytometric (FCM) minimalist strategy for TCD4+ enumeration. METHOD: 132 routine peripheral blood samples, mainly from HIV infected patients, were labelled with CD3-FITC/CD4-PE/CD45-PECy5 and analyzed by two gating methods: a) standard method based on Ly immunological gating (CD45++SSClow), followed by the determination of CD3+CD4+ percentage and absolute number (# calculation using Ly # from hematological analyser (HA); b) total WBC immunological gate on biparametric scatter CD45/CD4, followed by CD4++SSClow percentage determination and absolute number calculation using WBC absolute number from hematological counter without using the WBC differential. Moreover on 63 samples Ly # based on Ly % from FCM and WBC counting from HA was compared with Ly # from HA. RESULTS: The TCD4+/microL ranged from 3 to 3277 and the statistical analysis results showed: a) linear regression: r2 = 0.9847; b) Bland & Altman analysis: difference mean = -56.22; agreement range = +95.68 / -208.12; c) the mean of result difference/mean value*100 between two methods was -9.06%; d) comparison between regression line and the boundaries for acceptable residual values based on regressed confidence limits found by A. Kunkl et al showed regression line within boundaries near the upper limits. The Ly/microL count ranged from 635 to 8752. The statistical analysis results showed: a) linear regression: r2 = 0.9764; b) Bland & Altman analysis: difference mean = -362.93; agreement range = +134.51 / -860.37; c) the mean of result difference/mean value*100 between two methods was -16.12%. CONCLUSIONS: Our results suggest a fair agreement between the two gating methods, but the one based on total WBC gate gives TCD4+/microL counts systematically higher than the standard method. This finding can be attributed to a systematic lower estimation of Ly% by HA.

Antibodies, Monoclonal↗
PubMed2003

Pressure inactivation kinetics of Yersinia enterocolitica ATCC 35669.

The survival curves of Yersinia enterocolitica ATCC 35669 inactivated by high hydrostatic pressure were obtained at four pressure levels (300, 350, 400, and 450 MPa) in sodium phosphate buffer (0.1 M, pH 7.0) and four pressure levels (350, 400, 450, 500 MPa) in UHT whole milk. Tailing was observed in all the survival curves. A linear model and three nonlinear models were fitted to these data and the performances of these models were compared. The linear regression model for survival curves at four pressure levels had regression coefficients (R2) values of 0.785-0.962 and mean square error (MSE) of 0.265-0.893. A residual plot strongly suggested that a linear regression function was not appropriate as there was strong curvature in the plotted data. The nonlinear regression model using the log-logistic had R2 values of 0.946-0.982 and MSE values of 0.110-0.320. The Weibull model had R2 values of 0.944-0.975 and MSE values of 0.153-0.349. These results indicated that both were better models to describe the pressure inactivation kinetics of Y. enterocolitica in milk and buffer. Among the three nonlinear models studied, the modified Gompertz model produced the poorest fit to data. The number of parameters of the log-logistic model was reduced from four to two so that the model was greatly simplified. The reduced log-logistic model still produced a fit comparable to the full model. Since pressure had no significant effect on the shape factors of the Weibull model at the pressure levels of 300-400 MPa for buffer and 400-500 MPa for milk, models were developed to predict survival curves of Y. enterocolitica at pressures different from the experimental pressures.

Animals↗
PubMed2004

Applying an artificial neural network to warfarin maintenance dose prediction.

BACKGROUND: Oral anticoagulation with warfarin can lead to life-threatening events as a result of either over-anticoagulation or undertreatment. One of the main contributors to an undesirable warfarin effect is the need to adjust its daily dose for a specific patient. The dose is adjusted empirically based on the experience of the clinician, a method that is often imprecise. There is currently no other well-accepted method for predicting the maintenance dose of warfarin. OBJECTIVE: To describe the application of an artificial neural network to the problem of warfarin maintenance dose prediction. METHODS: We designed a neural network that predicts the maintenance dose of warfarin. Data on 148 patients attending a large anticoagulant clinic were collected by file review. Using correlational analysis of the patients' data we selected the best input variables. The network was trained by using the back-propagation algorithm on a subset of our data and the results were validated against the rest of the data. We used a multivariate linear regression to create a comparable model. RESULTS: The neural network generated reasonable predictions of the maintenance dose (r = 0.823). The results of the linear regression model were similar (r = 0.800). CONCLUSION: Neural networks can be applied successfully for warfarin maintenance dose prediction. The results are promising, but further investigation is needed.

Adolescent↗