Graphical Fit Research Articles

External validation of a paediatric Smart triage model for use in resource limited facilities.

Models for digital triage of sick children at emergency departments of hospitals in resource poor settings have been developed. However, prior to their adoption, external validation should be performed to ensure their generalizability. We externally validated a previously published nine-predictor paediatric triage model (Smart Triage) developed in Uganda using data from two hospitals in Kenya. Both discrimination and calibration were assessed, and recalibration was performed by optimizing the intercept for classifying patients into emergency, priority, or non-urgent categories based on low-risk and high-risk thresholds. A total of 2539 patients were eligible at Hospital 1 and 2464 at Hospital 2, and 5003 for both hospitals combined; admission rates were 8.9%, 4.5%, and 6.8%, respectively. The model showed good discrimination, with area under the receiver-operator curve (AUC) of 0.826, 0.784 and 0.821, respectively. The pre-calibrated model at a low-risk threshold of 8% achieved a sensitivity of 93% (95% confidence interval, (CI):89%-96%), 81% (CI:74%-88%), and 89% (CI:85%-92%), respectively, and at a high-risk threshold of 40%, the model achieved a specificity of 86% (CI:84%-87%), 96% (CI:95%-97%), and 91% (CI:90%-92%), respectively. Recalibration improved the graphical fit, but new risk thresholds were required to optimize sensitivity and specificity.The Smart Triage model showed good discrimination on external validation but required recalibration to improve the graphical fit of the calibration plot. There was no change in the order of prioritization of patients following recalibration in the respective triage categories. Recalibration required new site-specific risk thresholds that may not be needed if prioritization based on rank is all that is required. The Smart Triage model shows promise for wider application for use in triage for sick children in different settings.

Modeling the relationship between arterial blood pressure and sublingual microcirculatory blood flow assessed by Sidestream Dark Field videomicroscopy: An experimental study in anesthetized piglets.

Microcirculation is frequently assessed using videomicroscopy in the sublingual mucosa. However, limited research has been conducted on the existence of blood flow autoregulation in this region. We conducted a study in an experimental porcine model of pharmacologically induced hypotension to evaluate the relationship between mean arterial pressure (MAP) and the microvascular flow index (MFI). We hypothesized that this relationship would be linear in the absence of autoregulation or bilinear if autoregulation is present. Seven pigs underwent blood pressure changes induced by norepinephrine (hypertension) and sevoflurane (hypotension) administration. Sublingual microcirculation was assessed using a sidestream dark field device, and videos were recorded at different MAP levels ranging from 30 to 110 mmHg. MFI was calculated using the quadrant-based method. For our first hypothesis, we constructed a linear mixed model, while a bilinear model was used for the second hypothesis. The linear model demonstrated a statistically significant association (P = 0.03) described by the equation: MFI = 2.29 + 0.004 x MAP. The bilinear model identified a statistically significant inflection point at MAP = 99 mmHg (P = 0.01) with MFI = 2.7 AU (P < 0.0001). For MAP <99 mmHg, the relationship was: MFI = 2.26 + 0.004 x MAP, and for MAP >99 mmHg, MFI = 2.7. Despite statistical significance, neither model provided a satisfactory graphical fit due to high inter- and intra-individual variability. Consequently, this study did not allow us to draw conclusions regarding the presence of blood autoregulation.

Development and adaptation of relative sunshine hours models for the equatorial latitudes: a case study

This study focuses on the development, adaptation, and validation of relative sunshine hours (RSH) models, which are essential for localizing solar facilities in equatorial latitudes. The RSH models were developed from geographical and climatological variables on a periodic and non-periodic basis using 40-year (1979–2018) RSH data from the National Aeronautics and Space Administration (NASA). The NASA data were filtered and arranged in MS-Excel (V.16) to fit the structure of the developed RSH models and regressed in OriginLab to obtain the coefficients of the models. In the absence of RSH models, the clearness index models have been adapted (transformed) into RSH models. The measured data from the meteorological centers; Lira, Tororo, Kampala, and Mbarara were used in the validation of developed and adapted RSH models. Statistical tools; the root mean square error; RMSE yielded lower and upper limits (0.0635–0.2248), whereas the mean absolute percentage error, MAPE yielded a corresponding value (0.1620–2.5413). These results compare well with the literature results; RMSE (0.033–1.670) and MAPE (0.079–4.316), as the upper limits of the current results lie within the range of the literature results. In addition, the graphical fit of the empirical, developed, and adapted RSH data supports the robustness of the present RSH models. Validated RSH models are recommended for locating solar systems in equatorial latitudes. The oldest and adapted RSH models into equatorial latitudes have failed, so the present work further recommends for the redevelopment of such models with recent data to regenerate their validity.

Cost-effectiveness of first-line pembrolizumab plus chemotherapy for metastatic squamous non-small cell lung cancer (NSCLC).

e20703 Background: Trial results from KEYNOTE-407 have recently led to the FDA approval for pembrolizumab + carboplatin + paclitaxel/nab-paclitaxel (pembrolizumab+chemo) in previously untreated metastatic squamous NSCLC. This is the only first-line indication for squamous NSCLC regardless of tumor expression status. Our objective was to evaluate the cost-effectiveness of pembrolizumab combination therapy in this setting from the US payer perspective. Methods: Using data from KEYNOTE-407, we developed a partitioned survival decision model to estimate the lifetime costs and effectiveness of pembrolizumab+chemo vs. chemo alone in the first-line treatment of metastatic squamous NSCLC. The base case used a Weibull curve selected based on minimum AIC/BIC and best graphical fit to extrapolate in-trial survival to a lifetime horizon. First- and second-line therapy resource use and adverse event (AE) rates were derived from KEYNOTE-407. Utility data and AE management were obtained from published literature and national sources. Direct medical costs were adjusted to 2018 US dollars, and future costs and outcomes were discounted at 3% per year. We estimated life years (LY), quality-adjusted life years (QALYs), and costs over a lifetime horizon. One-way and probabilistic sensitivity analyses were also conducted. Results: In the base case, pembrolizumab+chemo resulted in 0.51 more LYs, 0.36 more QALYs, and $233,246 in healthcare costs vs. chemo alone. Costs per LY and QALY gained were $216,180 and $309,004, respectively. One-way sensitivity analyses indicated that the results were most sensitive to survival and pre-progression utility inputs. In a threshold analysis, we found that the cost of pembrolizumab+chemo would need to be reduced by 24% per course of therapy ($176,175) in order to be cost-effective at $150,000/QALY. Conclusions: Based on current available data, our analysis suggests first-line pembrolizumab-based combination therapy in metastatic squamous NSCLC is unlikely to be cost-effective relative to implied willingness to pay in cancer in the U.S. (ie &lt; $150,000 per QALY). Future studies should reassess cost-effectiveness as trial data mature.

Development of a rice tungro epidemiological model for seasonal disease risk management in the Philippines

Through an integration of seasonal climate forecasts (SCF) and rice disease epidemiological models, a potential risk for rice disease epidemics can be predicted even before a cropping season starts, allowing agricultural stakeholders to proactively manage risk on a seasonal scale. The objective of this study is to develop and evaluate an epidemiological “rtdSim” model for tungro, a vector-born virus disease of rice, aiming at producing a seasonal tungro risk prediction for the Bicol Region of the Philippines. Given the complex nature of disease pyramid (virus, vector, host, and environment) and their interactions with each other, a simplified model capable of predicting tungro epidemics makes use of assumption-based, simplified model algorithms. The rtdSim calculates the rate of rice hill infection with the rice tungro virus through its vector, the green leafhopper (GLH), using two linked modules: the “GLH population module” simulating complete GLH lifecycle population dynamics and immigration from nearby rice paddies; and the “Tungro virus infection module” simulating virus transmissions between infected rice hills and the GLH vector. The rtdSim is successfully calibrated and validated only for wet season, resulting in a reasonably high level of agreement: a good graphical fit for the incidence level of the disease at 60 days after transplanting and a temporal correlation coefficient of 0.70 between the observed and the simulated epidemics in the Bicol Region. The rtdSim simulations are sensitive to seasonal variations of temperature and rainfall, indicating the potential applicability of the model for seasonal RTD prediction with SCF. The present study highlights the potential for developing a SCF-based early warning system for rice diseases, enabling better decision-making for prevention of potential disease outbreaks on a seasonal scale with integrated pest management (IPM) strategies, e.g. switching to resistant varieties and using synchronous planting, rotations, or fallow periods to avoid peak risk pressure.

Shockley’s Equation Fit Analyses for Solar Cell Parameters from I-V Curves

Some of the technical problems that appear are obtaining solar cell parameters from I-V curve measurement data. One simple method is using linear graphical fit at zero current or voltage conditions. Although the accuracy of the obtained values is acceptable, other problems may arise regarding the number of parameters which could be obtained. We report a comparison between manual or graphical fit and fit using Shockley’s equation. The single I-V curve under the lighting was inferred to obtain the intrinsic parameters of the solar cells’ performance. The fittings were performed using the nonlinear equation of Shockley by determining some initial values of fittings such as Rs, Rsh, n, I0, Iph, and T. In the case of the Shockley equation fit, the iteration was performed several times to obtain the least possible inferred parameters. We have successfully obtained a better result of nonlinear Shockley fitting compared to the manual linear fit.

Distribution volume assessment compartment modelling: theoretic phosphate kinetics in steady state hemodialys patients.

Hyperphosphatemia constitutes a major problem in end-stage renal disease patients. At this stage, dialysis efficacy usually plays an important role in obtaining phosphate levels within the normal range. Currently, no practical tool capable of making individualized predictions about phosphate changes during and after hemodialysis (HD) have gained clinical acceptance. As a result, optimal dialysis prescriptions seem to be difficult to achieve. The objective of the present study was to develop and test a quantitative tool to predict intradialytic and postdialytic (2 hours) phosphate kinetics in HD therapy. This included distribution volume assessment. The approach included compartment modeling. Various model attempts were produced and tested using experimental data that included 2 treatment regimens; conventional and nocturnal HD, with 2-hour postdialysis rebound. Graphical comparison and determination of R(2) was applied to determine the best model variation. 1-, 2- and 3-compartment simulations were produced. Both 2- and 3-compartment model variations showed a close fit with the experimental data. However, a 3-compartment model showed the best graphical fit. This was supported by R(2) values in the 0.951-0.979 range. The 3-compartment model seems promising for prediction about plasma phosphate and holds the potential to be employed as a decision support tool and to enhance optimal dialysis prescriptions. Furthermore, the results provide specific suggestions about the distribution of phosphate in the body. Despite the promising results, further data and testing are necessary to validate the initial results.

Lateral spread of invasive Spartina alterniflora in uncrowded environments

We present a stochastic model of clonal growth in uncrowded environments and use it to study data of 7,536 clones from the invasion of Willapa Bay, Washington by the Atlantic cordgrass Spartina alterniflora. The model incorporates effects on clone growth of covariates, spatial autocorrelation, and temporal trends. The deterministic component of the model assumes that growth rate of a clone’s area is proportional to its perimeter, resulting in constant radial growth of the clone. The full stochastic model is built assuming that the fluctuations of radial growth increments (differences of square root-transformed areas) are normally distributed with constant variance. Graphical fit analysis with the density probability plot technique indicates that the stochastic model provides an excellent description of the S. alterniflora invasion. Variation in Spartina growth was significantly but weakly (5%) related to intertidal elevation, substrate type, year of survey, and the two-way interactions between these variables, suggesting that factors intrinsic to Spartina, along with localized high frequency noise, dominate the effects of larger scale abiotic factors on clone growth. Our model of clonal growth is potentially applicable to other systems with approximately circular plants or lichens.

Ureidosilanes on E-glass fibres: deposition and surface characteristics

The behaviour of methoxy- and ethoxy-ureidosilanes deposited onto E-glass fibres from both organic and aqueous solutions is reviewed. γ-Ureidopropyltri(m)ethoxysilanes are used as coupling agents for paints to metal surfaces, but they have not found much use as coupling agents for E-glass fibres. On metal surfaces, the silane is deposited with the ureido-functional group outermost, available for reaction with a paint/polymer matrix. On E-glass surfaces, however, unless specific conditions are employed, ureidosilanes tend to deposit with a highly cross-linked siloxane surface. The silanol groups condense not only with silanols of adjacent silanes and the E-glass surface, but also with aluminium hydroxides leached from the E-glass surface. This forms a highly cross-linked aluminosiloxane outer layer. The aluminosiloxane forms at all pHs between 4 and 10, and also when a range of organic deposition solvents is used. 29Si-NMR results are presented which illustrate the high level of condensation; silane to silicate binding via the amino group was not detected. A small proportion of Si—O—Si (silane-glass) bonds was inferred, indicating that a small number of sparsely distributed Si—O—Si links bound the silane to the surface. Electrokinetic analysis of the surfaces confirmed patchy deposition with a siloxane outer surface. Using AFM and ab initio calculations, it was estimated that the patchwise deposition of the silane was in the form of bilayers (or multiples of bilayers). The bilayer structure consisted of siloxane outer layers, with the inner ureido layers forming hydrogen bonds. A graphical fit using the mole and volume fractions of silane and organic solvents confirmed the various types of siloxane surfaces as deposited from different organic solvents. Chlorinated solvents tended to produce a ureido outer surface, which was condensed, but solventswollen; deposition from alcohol and acetone resulted in a siloxane surface. Toluene and THF both produced minimally condensed ureido surfaces, which were easily removed by washing with methylene chloride.

Application of radial basis function and feedforward artificial neural networks to the Escherichia coli fermentation process

Radial basis function and feedforward neural networks are considered for modelling of the recombinant Escherichia coli fermentation process. The models use industrial on-line data from the process as input variables in order to estimate the concentrations of biomass and recombinant protein, normally only available from off-line laboratory analysis. The models performances are compared by prediction error and graphical fit using results obtained from a common testing set of fermentation data.

Growth of research literature in scientific specialities. A modelling perspective

The paper discusses the application of three well known diffusion models and their modified versions to the growth of publication data in four selected fields of S&T. It is observed that all the three models in their modified versions generally improve their performance in terms of parameter values, fit statistics, and graphical fit to the data. The most appropriate model is generally seen to be the modified exponential-logistic model.

Data-Based Modelling Techniques for Reciprocal Space of NMR Imaging

In this paper the application of data-based modelling techniques to a reciprocal space components of a Nuclear Magnetic Resonance (NMR) irnaging system is considered. These include conventional identification techniques based on Auto-Regressive Moving Average with eXogenous inputs (ARMAX), Non-Linear ARMAX (NARMAX) and identifications based on Artificial Intelligence (AI) techniques such as Artificial Neural Networks (ANN). The mean sum of squared errors and graphical fit are used to compare the optimum performance of each technique

Type‐Curve Matching Using a Computer Spreadsheet

AbstractThis note shows how to construct and perform a graphical type‐curve analysis using a computer spreadsheet. Formulas are provided to generate the axis shifting behavior that is required for the graphical fit. Arbitrary match points are obtained by trial‐and‐error (or by closed form solution) for use in the calculation of aquifer hydraulic parameters. The analysis is illustrated using a leaky aquifer type curve and a confined aquifer type curve. Macro programming is not required, only data entry and the ability to use plotting features of a computer spreadsheet. The approach can be adapted for parameter estimation using any tabulated type curve.

What kind of materials are needed to build the structure of calculus?

The present article looks at another approach for presenting the ideas of calculus, instead of the traditional ‘limit based calculus’. Building the ideas of calculus from real applications, the first elements that should be studied are the ‘differential forms’. From these, the process of integration and the derivative appear naturally and are given meaning. Furthermore, the techniques of numerical and graphical integration fit very nicely within this approach. This paper discusses first, the different interpretations that can be associated with the three main elements of calculus: the differential, the derivative and the integral, and compares the approach suggested in this article with the conventional one. In doing this, it is shown how the Fundamental Theorem of Calculus can be interpreted as a ‘continuous sum’ of changes and it is shown why the geometric model of the integral as an area under the curve is a restricted view in many ways. The article then turns to examining the constructive materials of...

On data-based modelling techniques for fermentation processes

Six different modelling techniques were considered for the recombinant Escherichia coli fermentation process. These are Multiple Linear Regression (MLR), Principal Component Regression (PCR), Partial Least Squares (PLS), Auto-Regressive Moving Average with eXogeneous inputs (ARMAX), Non-linear ARMAX (NARMAX) and Artificial Neural Networks. The models use industrial on-line data from the process as input variables in order to forecast the concentrations of biomass and recombinant protein normally only available from off-line laboratory analysis. The models' performances are compared by prediction error and graphical fit using results obtained from a common testing set of fermentation data.

Comparing Different Modelling Techniques for the Escherichia Coli Fermentation Process

Four different modelling techniques are considered for the Escherichia coli fermentation process. These are Multiple Linear Regression (MLR), Principal Component Regression (TCR), Non-linear Auto-Regressive Moving Average with eXogencous inputs (NARMAX) and Artificial Neural Networks (ANNs). The models use industrial on-line data from the process as inputs in order to forecast the concentrations of biomass and recombinant protein, normally only available from offline laboratory analysis. The models’ performances are compared by prediction accuracy and graphical fit using results obtained from a common set of fermentation data.

On the Use of the Knox Equation. I. the Fit Problem

Abstract To obtain significant A, B, and C parameters of the Knox equation, which corresponds to the plot of the reduced plate height, h, versus the reduced linear velocity, v, actual results must be fitted correctly. The Knox equation is analyzed, the role of each individual parameter is shown. The equation producing the coordinates of the minimum plate height (maximum efficiency) is fully derived from the derivative of the Knox equation. Two tables giving the minimum coordinates for usual A, B, and C ranges are listed. The classical fit method is derived and analyzed. A graphical fit method which uses the remarkable graphical capabilities of modern spreadsheet packages is described. A synthetical set of data was fitted. It is demonstrated that each Knox parameter, obtained through fit procedures, must be given with the fitting confidence interval which is often in the 20% range.

Method of Evaluating Geometrical‐Statistical Parameters of the Interaction between Dislocations and Point Obstacles from Electron Micrographs

AbstractElectron micrographs of MgO crystals deformed in the HVEM show screw dislocations of curly shape. The dislocations are pinned at localized obstacles and bow out between them. The paper describes a measuring procedure for the geometrical interaction parameters between the dislocations and the obstacles. Such are the effective stress, the obstacle distance and the force acting on the obstacles. The procedure consists in a graphical fit of loops calculated by the DE WITT‐KOEHLER line tension model to the observed dislocation segments.

Human pharmacokinetics of sotalol.

Abstract The pharmacokinetics of sotalol were studied in healthy volunteers after a single dose and in hypertensive patients after chronic administration. A single 160 mg and two 80 mg socator® tablets were compared and found to be bioequivalent in terms of relative total bioavailability. Sotalol was well absorbed, about 73 % of the dose being excreted in the urine within 72 hours. The absolute bioavailability on oral administration was 100% indicating that both first–pass hepatic metabolism and metabolic destruction are negligible. There was no protein binding and the distribution volume was 171 1 after oral administration. The graphical fit of the plasma levels after a single oral dose was compatible with a two–compartment open model. The curve consisted of a distribution phase with a t1/2 of 6 hrs followed by a disposition phase with a t1/2 of 9.5 hrs. After chronic administration the t1/2 during the disposition phase was 10 hrs. Computer fitting of the data gave a terminal body half–life of 17.2 hrs which corresponded well with the terminal body half–life of 15 hrs calculated manually from urinary excretion rate data. In the computer analysis the steady–state plasma levels to be expected after long–term oral dosing were simulated and found to be in good agreement with experimental data, indicating that there is no induction of sotalol metabolism or increase in efficiency of excretion during long–term sotalol administration.

Exact phonon Green function for Kirkwood's model of polyethylene

The matrix elements of the lattice Green function for Kirkwood's model (describing the in-plane skeletal vibrations of an extended polyethylene chain) have been calculated exactly as a smooth function of frequency, using a semianalytic approach based on complex integration. The infinite set of matrix elements has been reduced to 12 basic real quantities from which any matrix element of the Green function in space-energy representation (including symmetry states) is obtained by linear combinations. Analytic expressions for these basic quantities are derived containing as parameter the solutions of a cubic equation. In principle, they could be evaluated on a desk calculator. Experimentally accessible frequencies are plotted as a function of the model parameters from which a graphical fit of the model can be easily found. The asymptotic behaviour of the Green function near these critical frequencies is studied in detail. As examples of straightforward applications, the density of states, the incoherent neutron scattering cross section and the localized modes due to an isotopic mass defect are plotted.