scholarly journals Fast computation of inverse transient analysis for pipeline condition assessment via surrogate modeling with sparse sampling strategy

2022 ◽  
Vol 162 ◽  
pp. 107995
Author(s):  
Xun Wang
Keyword(s):  
Transient Analysis ◽  
Surrogate Modeling ◽  
Sampling Strategy ◽  
Condition Assessment ◽  
Sparse Sampling ◽  
Fast Computation

scholarly journals Individualized Dosing of Hydroxyurea for Children with Sickle Cell Anemia Using a Population Pharmacokinetic-Based Model: The TREAT Study

Blood ◽  
2016 ◽  
Vol 128 (22) ◽  
pp. 3652-3652 ◽  
Author(s):  
Patrick T. McGann ◽  
Min Dong ◽  
Anu Marahatta ◽  
Thad A. Howard ◽  
Tomoyuki Mizuno ◽  
...  
Keyword(s):  
Dose Escalation ◽  
Research Funding ◽  
Sampling Strategy ◽  
Sparse Sampling ◽  
Primary Objective ◽  
Time Points ◽  
Individualized Dosing ◽  
Starting Dose ◽  
Population Pk ◽  
Sampling Approach

Abstract Background: Hydroxyurea is the primary disease modifying therapy for patients with sickle cell anemia (SCA). The clinical and laboratory benefits of hydroxyurea are the greatest when escalated to the maximum tolerated dose (MTD). The process of dose escalation to MTD requires expertise and can be tedious, often taking 6-12 months to titrate to the optimal dose. In addition, due in part to inter-patient variability in hydroxyurea pharmacokinetics (PK), the MTD varies among patients with a range of 15-35 mg/kg/day. We utilized a population PK model in combination with Bayesian estimation and a sparse sampling strategy, to individualize dosing of children starting hydroxyurea treatment. Methods: The Therapeutic Response Evaluation and Adherence Trial (TREAT, ClinicalTrials.gov NCT02286154) is a prospective study of hydroxyurea for children with SCA. The primary objective is to develop and evaluate a population PK-based model to predict hydroxyurea MTD through an individualized dosing strategy. A sparse sampling approach was developed to allow practical sampling from young children with SCA. The sampling strategy includes administering a single oral 20 mg/kg dose followed by collection of small quantities of blood (~100uL) at three post-dosing time points (15-20 minutes, 50-60 minutes, and ~3 hours). Baseline labs (including liver and renal function) are typically collected by venipuncture, while the other two samples are drawn by fingerstick or heelstick. Plasma hydroxyurea concentrations are measured using HPLC using an internal standard of methylurea. Using the population PK model with Bayesian estimation and hydroxyurea concentrations measured at the three specified time points, hydroxyurea exposure is estimated using specialized therapeutic drug monitoring software (MWPharm, Mediware, Prague, Czech Republic). Using the area under the curve (AUC0-inf) estimated by the model, we calculate a starting dose that is predicted to achieve an AUC of 115 ug*h/mL, which was the mean AUC value at MTD for a large cohort of children from a previous study (Dong M et al. Br J Clin Pharmacol 2016). The primary objective is to select a starting dose that is close to actual MTD, to reduce the time to maximum therapeutic effect and need for dose modifications before achieving MTD. Results: From December 2014 through June 2016, 20 children taking taking hydroxyurea for the first time were enrolled in TREAT. Seventeen of the 20 participants had all 3 post-treatment PK samples collected and processed to allow calculation of an individualized PK-based dose, while 3 had difficulties in sampling or processing that prevented a safe PK-guided dose recommendation. These 3 participants were started at the standard hydroxyurea dose of 20 mg/kg/day. The Table summarizes baseline characteristics for the initial study population, notable for a very young starting age with 13/20 (65%) less than two years of age. Twelve children with PK-based initial dosing have been treated with hydroxyurea for at least six months. Despite the young starting age, after six months of hydroxyurea, children have documented increases in total Hb (1.4+/-1.9 g/dL), HbF (11.3+/-6.4%), and MCV (15+/-8 fL) and decreases in absolute reticulocyte count (-217+/-128 x 109/L) and absolute neutrophil count (-1.0+/-1.9 x 109/L). In 9 of 12 participants, the PK-guided initial dose remained the best clinical dose at six months without significant dose changes except for minor adjustments for weight. Two patients required a single dose escalation due to inadequate marrow suppression, while one required a dose hold and decrease due to neutropenia during and following a viral infection. Conclusions: These data demonstrate that a sparse sampling approach, requiring only 3 blood samples over 3 hours, is able to accurately estimate hydroxyurea exposure in children with SCA. Hydroxyurea exposure, as defined by AUC, was similar with this sparse sampling approach as previous studies that relied upon a more standard and prolonged PK sampling approach. This population PK model is then able to predict a safe starting dose of hydroxyurea that approximates the actual MTD, with clinically significant improvements in laboratory parameters following six months of therapy. This individualized PK-guided dosing regimen should simplify hydroxyurea dosing and reduce the time interval to reach MTD and maximal clinical benefits. Table Table. Disclosures Kalfa: Baxter/Baxalta/Shire: Research Funding. Quinn:Silver Lake Research Corporation: Consultancy; Amgen: Research Funding; Eli Lilly: Research Funding. Ware:Nova Laboratories: Consultancy; Addmedica: Research Funding; Global Blood Therapeutics: Consultancy; Bayer Pharmaceuticals: Consultancy; Biomedomics: Research Funding; Bristol Myers Squibb: Research Funding.

scholarly journals Sensor Placement Strategy for Pipeline Condition Assessment Using Inverse Transient Analysis

2019 ◽  
Vol 33 (8) ◽  
pp. 2761-2774 ◽  
Author(s):  
Chi Zhang ◽  
Jinzhe Gong ◽  
Martin F. Lambert ◽  
Angus R. Simpson ◽  
Aaron C. Zecchin
Keyword(s):  
Transient Analysis ◽  
Sensor Placement ◽  
Condition Assessment

A Rational Design Approach to Gaussian Process Modeling for Variable Fidelity Models

2011 ◽  
Author(s):  
Roxanne A. Moore ◽  
David A. Romero ◽  
Christiaan J. J. Paredis
Keyword(s):  
Gaussian Process ◽  
Surrogate Model ◽  
Rational Design ◽  
Sequential Sampling ◽  
Accurate Determination ◽  
Surrogate Modeling ◽  
Sampling Strategy ◽  
New Approach ◽  
Variable Fidelity ◽  
The Cost

Computer models and simulations are essential system design tools that allow for improved decision making and cost reductions during all phases of the design process. However, the most accurate models tend to be computationally expensive and can therefore only be used sporadically. Consequently, designers are often forced to choose between exploring many design alternatives with less accurate, inexpensive models and evaluating fewer alternatives with the most accurate models. To achieve both broad exploration of the design space and accurate determination of the best alternatives, surrogate modeling and variable accuracy modeling are gaining in popularity. A surrogate model is a mathematically tractable approximation of a more expensive model based on a limited sampling of that model. Variable accuracy modeling involves a collection of different models of the same system with different accuracies and computational costs. We hypothesize that designers can determine the best solutions more efficiently using surrogate and variable accuracy models. This hypothesis is based on the observation that very poor solutions can be eliminated inexpensively by using only less accurate models. The most accurate models are then reserved for discerning the best solution from the set of good solutions. In this paper, a new approach for global optimization is introduced, which uses variable accuracy models in conjuction with a kriging surrogate model and a sequential sampling strategy based on a Value of Information (VOI) metric. There are two main contributions. The first is a novel surrogate modeling method that accommodates data from any number of different models of varying accuracy and cost. The proposed surrogate model is Gaussian process-based, much like classic kriging modeling approaches. However, in this new approach, the error between the model output and the unknown truth (the real world process) is explicitly accounted for. When variable accuracy data is used, the resulting response surface does not interpolate the data points but provides an approximate fit giving the most weight to the most accurate data. The second contribution is a new method for sequential sampling. Information from the current surrogate model is combined with the underlying variable accuracy models’ cost and accuracy to determine where best to sample next using the VOI metric. This metric is used to mathematically determine where next to sample and with which model. In this manner, the cost of further analysis is explicitly taken into account during the optimization process.

scholarly journals On-site non-invasive condition assessment for cement mortar–lined metallic pipelines by time-domain fluid transient analysis

2015 ◽  
Vol 14 (5) ◽  
pp. 426-438 ◽  
Author(s):  
Jinzhe Gong ◽  
Mark L Stephens ◽  
Nicole S Arbon ◽  
Aaron C Zecchin ◽  
Martin F Lambert ◽  
...  
Keyword(s):  
Time Domain ◽  
Transient Analysis ◽  
Cement Mortar ◽  
Condition Assessment ◽  
Non Invasive ◽  
Fluid Transient

scholarly journals Simulation of Brisk and Fast Phase-Contrast Magnetic Resonance Imaging by Computational Fluid Dynamics

2003 ◽  
Author(s):  
Bradley L. Hershey ◽  
Mark Doyle ◽  
Eduardo Kortright ◽  
Andreas Anayiotos
Keyword(s):  
Magnetic Resonance Imaging ◽  
Fluid Dynamics ◽  
Computational Fluid Dynamics ◽  
Magnetic Resonance ◽  
Phase Contrast ◽  
Sampling Rate ◽  
Sampling Strategy ◽  
Sparse Sampling ◽  
Resonance Imaging ◽  
Sampling Schemes

Cardiac synchronized magnetic resonance imaging of flowfields has suffered due to the relatively long acquisition times required. We developed a rapid MRI approach, BRISK PCA (Block Regional Interpolation Scheme for k-space Phase Contrast Angiography) which was simulated here using data generated by computational fluid dynamics to investigate the role of interpolation and segmentation on the accuracy and efficiency of the method. BRISK differs from other sparse sampling schemes in that the sampling rate is a function of the position in k-space and interpolation is used to generate data points not directly acquired. Combined with conventional segementation, this allows more efficient use of time, resulting in rapid acquisitions with good spatial and temporal resolution. FAST (Fourier AcquiSition in Time) is a similar sparse sampling strategy that varies the segmentation factor, rather than the sampling rate, as a function of k-space position. BRISK and FAST can be performed in nearly equally scan times. However, deviation from ideal in the FAST data was highly dependant on the starting phase of the flow waveform, while BRISK was immune to such variation. Simulations showed that BRISK (up to segmentation factor 5) and FAST 5 retained excellent axial-velocity accuracy, but the accuracy of FAST was variable and dependent on waveform characteristics.

scholarly journals Multi-stage parameter-constraining inverse transient analysis for pipeline condition assessment

2018 ◽  
Vol 20 (2) ◽  
pp. 281-300 ◽  
Author(s):  
Chi Zhang ◽  
Aaron C. Zecchin ◽  
Martin F. Lambert ◽  
Jinzhe Gong ◽  
Angus R. Simpson
Keyword(s):  
Transient Analysis ◽  
Distribution Systems ◽  
Water Distribution ◽  
Search Space ◽  
Condition Assessment ◽  
Estimation Accuracy ◽  
Model Parameters ◽  
Multi Stage ◽  
Study Results ◽  
Transmission Pipelines

Abstract Fault detection in water distribution systems is of critical importance for water authorities to maintain pipeline assets effectively. This paper develops an improved inverse transient analysis (ITA) method for the condition assessment of water transmission pipelines. For long transmission pipelines ITA approaches involve models using hundreds of discretized pipe reaches (therefore hundreds of model parameters). As such, these methods struggle to accurately and uniquely determine the many parameter values, despite achieving a very good fit between the model predictions and measured pressure responses. In order to improve the parameter estimation accuracy of ITA applied to these high dimensional problems, a multi-stage parameter-constraining ITA approach for pipeline condition assessment is proposed. The proposed algorithm involves the staged constraining of the parameter search-space to focus the inverse analysis on pipeline sections that have a higher likelihood of being in an anomalous state. The proposed method is verified by numerical simulations, where the results confirm that the parameters estimated by the proposed method are more accurate than the conventional ITA. The proposed method is also verified by a field case study. Results show that anomalies detected by the proposed methods are generally consistent with anomalies detected by ultrasonic measurement of pipe wall thickness.

Sign in / Sign up

Export Citation Format

Share Document