Surrogate Response Research Articles

Sufficient dimension reduction via distance covariance for survival data

Sufficient dimension reduction (SDR) plays an essential role in high-dimensional regression analysis, and most existing SDR methods suffer when data are subject to censoring. This paper tries to extend distance covariance (dCov) to SDR for survival data. Unlike most methods requiring strict constraints on predictors and survival time, the method based on dCov only relies on very mild conditions and works well when many predictors are categorical or discrete. The difficulty of SDR for survival data is that we cannot completely observe the real survival time, which may introduce a substantial bias. When the censoring time is independent of survival time, we show that the dCov-SDR method can estimate the SDR directions with a surrogate response. Furthermore, under other censoring assumptions, we extend the dCov-SDR method by proposing a new consistent empirical version of dCov for survival data. We discuss the inherent properties of our method, and under regularity conditions, statistical consistency is established for our estimators. Both simulations and real data analysis demonstrate the promising performance of the proposed method.

A multi-fidelity approach for the evaluation of extreme wave loads using nonlinear response-conditioned waves

In the present study, a multi-fidelity methodology based on response-conditioned waves (RCW) is demonstrated for the probabilistic assessment of wave-induced loads and responses of offshore structures. The methodology consists of two steps: (i) the RCW are determined using a surrogate response model and (ii) they are reproduced in an experimental or CFD-based numerical wave tank (NWT) to obtain the fully nonlinear response, as a high-fidelity evaluation. The main novelty of the proposed response-conditioning technique is that it uses a fully nonlinear wave solver that calculates the wave propagation inside an NWT. In this work, the extreme Vertical Bending Moment (VBM) of a zero-speed containership is investigated. It is found that the proposed approach provides nonlinear wave sequences that can be explicitly and exactly reproduced experimentally, up to very extreme sea states (Hs=17 m, Tp=15.5 s). Moreover, through comparison of the obtained results with an experimental Monte Carlo approach, it is shown that the overall framework can predict the short-term distribution of the VBM accurately and efficiently. Finally, given that model tests are costly and not widely accessible, the implementation of the high-fidelity response evaluation within a CFD-based NWT is also demonstrated and validated against the available experimental results.

Exploratory analysis of serum HER2 extracellular domain for HER2 positive gastric cancer treated with SOX plus trastuzumab.

The aim of this study was to explore the clinical utility of serum HER2 extracellular domain (sHER2 ECD) using data from a clinical trial evaluating trastuzumab combined S-1 plus oxaliplatin (SOX) in HER2 positive gastric cancer. sHER2 ECD were prospectively measured at baseline and subsequent treatment courses. Based on each quantile point of baseline sHER2 ECD levels and its early changes, patients were divided into two groups and compared clinical outcomes. 43 patients were enrolled, and 17 patients (39.5%) were positive for baseline sHER2 ECD. Higher baseline sHER2 ECD levels tended to have lower hazard ratios (HRs). When divided into two groups by baseline sHER2 ECD of 19.1ng/ml, median progression-free survival (PFS) and overall survival (OS) was longer in the higher group (mPFS: 16.8 vs 8.7months, p = 0.359. mOS: 35.5 vs 20.6months, p = 0.270), respectively. After initiation of treatment, sHER2 ECD significantly decreased up until the third cycle. Higher reduction rates of sHER2 ECD within 3 cycles also tended to have lower HRs. When divided into two groups by reduction rate of 42.5%, mPFS and mOS was longer in the higher reduced group (mPFS: 17.2 vs 8.7months, p = 0.095. mOS: 65.0 vs 17.8months, p = 0.047), respectively. Furthermore, higher reduction rates could surrogate higher objective response rates (ORR) (ORR: 90% vs 63.2% for 29.5%, p = 0.065. 100% vs 70% for 42.5%, p = 0.085), respectively. Baseline sHER2 ECD levels and its early decline may be useful biomarkers for SOX plus trastuzumab efficacy in HER2 positive gastric cancer.

Bilayer surrogate brain response under various blast loading conditions

Variations in the experimental constraints applied within blast simulations can result in dramatically different measured biomechanical responses. Ultimately, this limits the comparison of data between research groups and leads to further inquisitions about the “correct” biomechanics experienced in blast environments. A novel bilayer surrogate brain was exposed to blast waves generated from advanced blast simulators (ABSs) where detonation source, boundary conditions, and ABS geometry were varied. The surrogate was comprised of Sylgard 527 (1:1) as a gray matter simulant and Sylgard 527 (1:1.2) as a white matter simulant. The intracranial pressure response of this surrogate brain was measured in the frontal region under primary blast loading while suspended in a polyurethane spherical shell with 5 mm thickness and filled with water to represent the cerebrospinal fluid. Outcomes of this work discuss considerations for future experimental designs and aim to address sources of variability confounding interpretation of biomechanical responses.

Biomechanical Response of Head Surrogate With and Without the Helmet.

Measurements of brain deformations under injurious loading scenarios are actively sought. In this work, we report experimentally measured head kinematics and corresponding dynamic, two-dimensional brain simulant deformations in head surrogates under a blunt impact, with and without a helmet. Head surrogates used in this work consisted of skin, skull, dura, falx, tentorium, and brain stimulants. The head surrogate geometry was based on the global human body models consortium's head model. A base head surrogate consisting of skin-skull-brain was considered. In addition, the response of two other head surrogates, skin-skull-dura-brain, and skin-skull-dura-brain-falx-tentorium, was investigated. Head surrogate response was studied for sagittal and coronal plane rotations for impactor velocities of 1 and 3 m/s. Response of head surrogates was compared against strain measurements in PMHS. The strain pattern in the brain simulant was heterogenous, and peak strains were established within ∼30 ms. The choice of head surrogate affect the spatiotemporal evolution of strain. For no helmet case, peak MPS of ∼50-60% and peak MSS of ∼35-50% were seen in brain simulant corresponding to peak rotational accelerations of ∼5000-7000 rad/s2. Peak head kinematics and peak MPS have been reduced by up to 75% and 45%, respectively, with the conventional helmet and by up to 90% and 85%, respectively, with the helmet with antirotational pads. Overall, these results provide important, new data on brain simulant strains under a variety of loading scenarios-with and without the helmets.

Correlating GC/MS Relative Response Factors to Analyte's Physicochemical and Chromatographic Properties to Facilitate the Quantitation of Organic Extractables and Leachables in Non-Targeted Analysis (NTA). Concepts and Empirical Considerations.

Leachables in drug products and from medical devices can adversely affect patient health and thus must be identified and quantified. Accurate and protective quantitation in target analysis for leachables (and extractables as potential leachables) is accomplished via compound-specific calibration curves. Quantification in non-targeted analysis (NTA) is complicated by the variable relative response factors (RRFs) among and between individual leachables and the circumstance that the leachables are not known until the NTA is completed. Protective quantitation in NTA is accomplished in various ways, depending on the identification status of the analyte. When an analyte's identity is confirmed, it is quantified using the compound's own RRF, obtained by analysis of a reference standard. In other identification circumstances, the concentration is calculated using a surrogate response, either linked to a surrogate compound or representative of a domain of leachables. Given the difficulty in matching an analyte with a proper surrogate, this article addresses quantitation via the latter approach. This article uses a database of >3000 GC/MS response factors to empirically divide the population of leachables into response factor domains, differentiated by either the analyte's polarity (log Po/w) or retention time. Using the database, mean RRF values and uncertainty factors (UFs) are established for each domain and are used for quantitation. Protective quantitation is accomplished for nearly 84% of all leachables in the database (and presumably the entire population) by placing an analyte into its proper domain and then using the mean RRF divided by the UF for that domain as a universal response factor for all compounds in the domain.

Conditional density function for surrogate scalar response

his paper presents the estimator of the conditional density function of surrogated scalar response variable given a functional random one. We construct a conditional density function by using the available (true) response data and the surrogate data. Then, we build up some asymptotic properties of the constructed estimator in terms of the almost complete convergences. As a result, we compare our estimator with the classical estimator through the Relatif Mean Square Errors (RMSE). Finally, we end this analysis by displaying the superiority of our estimator in terms of prediction when we are lacking complete data.

Study on impact resistance and parameter optimization of patch‐repaired plain woven composite based on multi‐scale analysis

AbstractPatch‐repaired plain woven composite structures were exposed to internal damage caused by impact during service, which reduced mechanical properties. A multi‐scale modeling strategy was proposed to study the impact resistance characteristics of patch‐repaired plain woven carbon‐fiber‐reinforced‐polymer (CFRP) laminates. The micro representative volume element (RVE) was extracted from the yarn; a local homogenization method was used to obtain the equivalent cross‐ply laminate (ECPL) cells corresponding to the mesoscale RVE. The finite element model (FEM) of patch‐repaired plain woven laminate was established by extending the ECPL cells. The FEM was checked by the low‐velocity impact (LVI) tests, and the impact parameter errors were within 10%. Furthermore, impact resistance of repaired specimens under different impact conditions (e.g., the impact energy was 3 ~ 20 J) was studied. Finally, patch parameters (patch thickness, patch size, off‐axis angle) of the repaired specimen were optimized by multi‐objective optimization based on the surrogate model and response surface method (RSM). After multi‐objective function optimization, the optimal patch parameter configuration (the thickness of the patch was 0.87 mm, the diameter was 66 mm, and the off‐axis angle was 44.2°) reduced the absorbed energy value of the repaired specimen from 4.9089 to 4.1640 J, and the delamination area of the repaired specimen from 605 to 280 mm2. All the results provide a reference for patch repair of plain woven composites.

Generalized regression function for surrogate scalar response

In this paper we develop and generalize the estimator of regression function for surrogate scalar response variable given a functional random one. Then, we build up some asymptotic properties in terms of the almost complete convergences, depending in the result we show the superiority of our estimator in term of prediction.

Dynamic contrast-enhanced MRI parameters and apparent diffusion coefficient as treatment response markers of skull base osteomyelitis: a preliminary study.

Currently, there is no definitive consensus on the optimal imaging modality for determining the treatment response in patients with skull base osteomyelitis (SBO). This study aimed to investigate the utility of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) parameters and apparent diffusion coefficient (ADC) as treatment response markers of SBO. This study included 6 patients with SBO, who underwent both pre- and post-treatment DCE-MRI and diffusion-weighted imaging (DWI). Quantitative DCE-MRI parameters and ADC of the region-of-interest were analysed. These normalized parameters were calculated by dividing the region-of-interest by the reference region. The Wilcoxon signed rank test was used to compare these parameters between pre- and post-treatment periods. The normalized fraction of the extravascular extracellular space (Ve) and ADC of the post-treatment status of SBO was significantly lower than those of pre-treatment measures (p = 0.03). The normalized fraction of blood plasma (Vp), normalized rate of transfer from the blood plasma into the extravascular extracellular space (Ktrans), and normalized backflow leakage of material from the extravascular extracellular space into the blood plasma (Kep) demonstrated no significant differences between pre- and post-treatment. DCE-MRI parameters Ve and ADC demonstrated a significant reduction when comparing measures across the pre- and post-treatment periods. These parameters may potentially serve as a valuable surrogate treatment response marker for SBO activity.

Crashworthiness analysis and multiobjective optimization for variable thickness square thin-wall columns under axial loading

This article introduces a novel square cross-section thin wall column (TWC) to enhance crashworthiness under axial loading. The wall thickness of the square cross-section TWC was symmetrically designed to increase crashworthiness. This type of TWC is known as variable thickness (VT) TWC. Crashworthiness performance was measured using indicators such as energy absorption (EA), specific energy absorption (SEA), and crash force efficiency (CFE). The uniform thickness (UT) cross-section TWC was employed for experimental testing and analytical evaluation, and its crashworthiness performance was considered a benchmark for this study. The experimental test was simulated in ABAQUS explicit dynamic software, and experimental, numerical, and analytical results were all compared. The same simulation model was used for subsequent investigations on VT-TWC. In order to find the most effective VT-TWC, which has the highest SEA and CFE indicators, the multi-objective optimization problem was formulated using a surrogate model and response surface technique. The Non-dominated Sorting Genetic Algorithm (NSGA-II) was employed to solve the multi-objective problem for maximized SEA and CFE. From the results, it was predicted that the VT-TWC performs significantly better than the UT-TWC, saving around 15.63% of the material with 0.45% and 9.5% higher SEA and CFE, respectively.

A new goodness-of-fit measure for probit models: Surrogate R2.

Probit models are used extensively for inferential purposes in the social sciences as discrete data are prevalent in a vast body of social studies. Among many accompanying model inference problems, a critical question remains unsettled: how to develop a goodness-of-fit measure that resembles the ordinary least square (OLS) R2 used for linear models. Such a measure has long been sought to achieve 'comparability' of different empirical models across multiple samples addressing similar social questions. To this end, we propose a novel R2 measure for probit models using the notion of surrogacy - simulating a continuous variable as a surrogate of the original discrete response (Liu & Zhang, Journal of the American Statistical Association, 113, 845 and 2018). The proposed R2 is the proportion of the variance of the surrogate response explained by explanatory variables through a linear model, and we call it a surrogate R2 . This paper shows both theoretically and numerically that the surrogate R2 approximates the OLS R2 based on the latent continuous variable, preserves the interpretation of explained variation, and maintains monotonicity between nested models. As no other pseudo R2 , McKelvey and Zavoina's and McFadden's included, can meet all the three criteria simultaneously, our measure fills this crucial void in probit model inference.

Sensitivity analysis for uncertain periodically excited nonlinear systems through kriging based harmonic balance method

There exist various indeterministic dynamical systems in practical applications due to various uncertainties in structural geometry, material and loading. It is generally difficult to acquire a priori probabilistic characteristics of these uncertainties based on onsite data later to be used in random dynamical assessments. Therefore, non-probabilistic methods such as interval analysis are suitable for systems carrying uncertain but bounded parameters. Interval analysis method generally resorts to Taylor series expansions of intervallic parameters for estimation of lower and upper bounds of dynamical responses which are solved by perturbation finite element method (PFEM). Nevertheless, computational error is inevitably accumulated during temporal integration of the responses as well as evaluation of their derivatives with respect to parameters. To circumvent this drawback, an improved interval analysis approach called Kriging-surrogate-based general harmonic balance method (KGHBM) is proposed in this paper. In this method, harmonics in the Fourier expansion of the dynamical response are solved by the generalized harmonic balance method in the frequency domain. A surrogate response surface of the harmonic coefficients is established in the parameter space for evaluating the sensitivities of the response with respect to the parameters. Compared with the PFEM, the present method provides better bound predictions of uncertain nonlinear dynamical response. KGHBM can be modified with the adoption of second-order derivatives in the Taylor expansion on the Kriging surrogate model to improve the accuracy of the KGHBM predictions.

Deterministic and probabilistic prediction of maximum wall facing displacement of geosynthetic-reinforced soil segmental walls using multivariate adaptive regression splines

A surrogate model-based response surface method is proposed to predict the maximum wall facing displacement of geosynthetic-reinforced soil (GRS) segmental walls using the multivariate adaptive regression splines (MARS). The synthetic datasets are generated using the Latin hypercube sampling (LHS) technique for both training and testing purposes. For each input dataset, the maximum wall facing displacement is evaluated using the finite difference software FLAC. An expression for the maximum wall facing displacement is proposed in terms of basis functions (piecewise linear). The performance of the proposed model is assessed in terms of the ratio of predicted and simulated maximum wall facing displacement, and the results show that the performance is satisfactory. Moreover, the study identifies that the effect of soil friction angle is most apparent in predicting the maximum wall facing displacement. The efficiency of the proposed model is compared with other soft computing techniques, and the result indicates that the proposed model gives minimum error. Further, a probabilistic study is performed in terms of the normalized maximum wall facing displacement. The results show that the probabilistic prediction of the maximum wall facing displacement using the proposed model is within the range of specified limit.

Proteomic analysis of plasma exosomes as biomarkers of response to MVP-S based immunotherapy

Abstract Ovarian cancer is a deadly gynaecological disease owing to its late-stage diagnosis. Recurrence and chemo-resistance develop despite initial treatment success. Maveropepimut-S (MVP-S), formerly DPX-Survivac, is a DPX-based immune educating therapy that elicits robust, targeted and, sustained specific anti-tumor T- and B-cell responses in advanced ovarian cancer patients. Exosomes are a subtype of extracellular vesicles (EVs) secreted by cells involved in intercellular communication. They contain bioactive molecules that can influence the tumor, the immune system, and the extracellular environment. Hence, circulating exosomes could provide real time assessment of disease evolution making them ideal minimally invasive biomarkers for monitoring response to immunotherapies. Plasma exosomes were isolated from advanced ovarian cancer patients (n=22) at pre- and on-treatment with MVP-S based therapy (DeCidE1 trial, NCT02785250). Subjects with clinical benefit (>10% tumor shrinkage) to therapy displayed a lower overall total EV protein concentration. Untargeted proteomic profiling by liquid chromatography tandem mass spectrometry identified 227 proteins in exosomes. Of these, 95 (42%) were likely from non-tumor exosomes and involved in immune modulation such as regulation of humoral response and B-cell immunity. When compared to the plasma proteome, 128 (56%) proteins were found to be exclusive to exosomes. Functional annotations of these proteins suggest their role in modulating immune pathways related to Fc receptor mediated signaling and T-cell migration. The results highlight the potential value of plasma exosome assays for monitoring disease and as a potential surrogate response biomarker. Supported by -IMV inc. -NRC-IRAP R&D Certificate Program.

Fatigue Optimization of Structural Parameters for Orthotropic Steel Bridge Decks Using RSM and NSGA-II

This study presents an optimization flow for structural parameters of orthotropic steel bridge decks (OSDs) based on the fatigue performance and the structural weight. The effects of structural parameters, namely, the thicknesses of the deck plate, the U-rib, the diaphragm, and the longitudinal diaphragm, on equivalent fatigue stress ranges were investigated using surrogate modelling based on the response surface methodology (RSM). Finally, the Pareto optimum front was obtained via the nondominated sorting genetic algorithm II (NSGA-II). The results show that the established surrogate response surface model can predict the local equivalent stress amplitudes well. The deck thickness is the most significant factor for the fatigue performance of the deck to U-rib weld and the U-rib to diaphragm weld. The objectives of minimizing the equivalent stress range and the total structural weight are truly conflicting, and the optimum solutions for structural parameters can be obtained from the Pareto optimum front.

A Study on the Optimization of In-Process Inspection Procedure for Active Pharmaceutical Ingredients Manufacturing Process

The in-process inspection procedure is one of the critical operations in the active pharmaceutical ingredients manufacturing process. This study aims to improve the performance of the IPI service system in terms of three main criteria, namely service level, cycle time, and maximum tardy time. In solving this multiple-criteria decision-making problem, the proposed study seeks to redesign three process control factors, namely the service configuration, the dispatching rule, and the scheduling rule. The problem is solved using the Taguchi robust design methodology. Since the Taguchi method handles parameter design problems with only one criterion, Technique for Order Preference by Similarity to an Ideal Solution, a multiple-criteria decision-making method, is used to provide a surrogate response to the Taguchi method. The numerical results show that the redesigned IPI system improves the service level by 28.75%, the cycle time by 18.32%, and the maximum tardy time by 22.22%.

Global optimization on non-convex two-way interaction truncated linear multivariate adaptive regression splines using mixed integer quadratic programming

Multivariate adaptive regression splines (MARS) has been adopted as a popular surrogate function to model the unknown relationships between input and output variables in complex systems. Searching optimal solutions on the complicated surrogate response surface of MARS serves as important tasks in various applications. In this paper, we present an efficient and effective approach to find a global optimal value of MARS models that incorporate two-way interaction terms which are products of truncated linear univariate functions (TITL-MARS). Specifically, with a MARS model consisting of linear and quadratic structures, we can reformulate the optimization problem on TITL-MARS into a mixed integer quadratic programming problem (TITL-MARS-OPT), which can be further solved in a more principled way. To illustrate the effectiveness of our proposed approach TITL-MARS-OPT, we come up with a genetic algorithm and a gradient descent algorithm to solve the original version of TITL-MARS, and we compared the performance of the proposed approach with the benchmark algorithms. Numerical experiments are conducted on a spectrum of examples with different levels of complexity, including 6 existing models and a real world application in the wind farm optimization. Our proposed approach can find a global optimal successfully and efficiently while the comparison algorithms fail to find a global optimal solution. In the end, Python code for TITL-MARS and TITL-MARS-OPT is made available on GitHub ( https://github.com/JuXinglong/TITL-MARS-OPT).

On the Potential of Sequential and Nonsequential Regression Models for Sentinel-1-Based Biomass Prediction in Tanzanian Miombo Forests

This study derives regression models for above-ground biomass (AGB) estimation in miombo woodlands of Tanzania that utilise the high availability and low cost of Sentinel-1 data. The limited forest canopy penetration of C-band SAR sensors along with the sparseness of available ground truth restrict their usefulness in traditional AGB regression models. Therefore, we propose to use AGB predictions based on airborne laser scanning (ALS) data as a surrogate response variable for SAR data. This dramatically increases the available training data and opens for flexible regression models that capture fine-scale AGB dynamics. This becomes a sequential modelling approach, where the first regression stage has linked in situ data to ALS data and produced the AGB prediction map; We perform the subsequent stage, where this map is related to Sentinel-1 data. We develop a traditional, parametric regression model and alternative non-parametric models for this stage. The latter uses a conditional generative adversarial network (cGAN) to translate Sentinel-1 images into ALS-based AGB prediction maps. The convolution filters in the neural networks make them contextual. We compare the sequential models to traditional, non-sequential regression models, all trained on limited AGB ground reference data. Results show that our newly proposed non-sequential Sentinel-1-based regression model performs better quantitatively than the sequential models, but achieves less sensitivity to fine-scale AGB dynamics. The contextual cGAN-based sequential models best reproduce the distribution of ALS-based AGB predictions. They also reach a lower RMSE against in situ AGB data than the parametric sequential model, indicating a potential for further development.

Evaluation of pathological complete response as surrogate endpoint in neoadjuvant randomised clinical trials of early stage breast cancer: systematic review and meta-analysis

ObjectiveTo evaluate pathological complete response as a surrogate endpoint for disease-free survival and overall survival in regulatory neoadjuvant trials of early stage breast cancer.DesignSystematic review and meta-analysis.Data sourcesMedline, Embase, and...