Correlated Model Parameters Research Articles

Optimal sampling strategy for probability estimation: An application to the Agricultural Quarantine Inspection Monitoring program

AbstractImported agricultural pests can cause substantial damage to agriculture, food security, and ecosystems. In the United States, the Agricultural Quarantine Inspection Monitoring (AQIM) program conducts random sampling to estimate the probabilities that cargo and passengers arriving at ports of entry carry pests. Assessing these risks accurately is critical to enable effective policies and operational procedures. This study introduces a pathway‐level analysis with an objective function aligned with AQIM's goal, offering a new perspective compared to the current container‐by‐container approach, which relies on heuristics to set inspection rates. We formulate an optimization model that minimizes the mean squared error of the probability estimates that AQIM obtains. The central decision‐making tradeoff that the model explores is whether it is preferable to sample more arriving containers (and fewer boxes per container) or more boxes per container (and fewer containers), given limited resources. We first derive an analytical solution for the optimal sampling strategy by leveraging several approximations. Then, we apply our model to a numerical case study of maritime cargo sampling at the Port of Long Beach. Across a wide range of parameter settings, the optimal strategy samples more containers (but fewer boxes per container) than the current AQIM protocol. The difference between the two strategies and the accuracy improvement with the optimal approach are larger if the pest statuses of boxes in the same container are more strongly correlated. We recommend that AQIM record box‐level (beyond only container‐level) inspection data, which could be used to estimate this correlation and other model parameters.

Surface integrity analysis and inspection for nanochannel sidewalls using the self-affine fractal model-based statistical quality control for the atomic force microscopy (AFM)-based nanomachining process

The atomic force microscopy (AFM) technology is a promising method for nanofabrication due to the high tunability of this affordable platform. The quality inspection and control significantly impact the manufacturing effectiveness for realizing the functionality of the achieved nanochannel. Particularly, the surface characteristics of nanochannel sidewalls, which play a significant role in determining the quality of the nanomachined products, can not be accurately captured using conventional surface integrity metrics (e.g., surface roughness). Therefore, it is necessary to propose a method to quantitatively characterize the surface morphology and detect the abnormal parts/regions of the nanochannel sidewall. This paper presents a statistical process control approach derived from the self-affine fractal model to detect the sidewall surface anomalies. It evaluates changes in the self-affine fractal model parameters (standard deviation, correlation length, and roughness exponent), which can be used to signify the changes on the sidewall surface; the statistical distributions of these parameters are derived and used to develop control charts to allow inspection of the sidewall morphology. The approach was tested on the AFM-based nanomachined samples. The results suggest that the presented approach can effectively reflect the abnormal regions on the machined parts, which opens up a new avenue toward guiding the quality control and rework for process improvement for AFM-based nanomachining.

Relationship between topographic parameters and BRDF for tungsten surfaces in the visible spectrum

In metallic fusion devices, parasitic light originating from multiple reflections on the wall is a major problem for the interpretation of optical diagnostics. Strong stray light affects several optical diagnostics in ITER. One possibility to cope with this reflected light is to use photonic simulation, which can accurately predict the behavior of light within complex 3D geometry. A prerequisite is to get a good description of the reflection model, represented by the Bidirectional Reflectance Distribution Function (BRDF), based on optical measurements of in-vessel materials. To avoid complicated measurements using goniophotometer to get the BRDF, one possibility is to link surface optical properties and topography characteristics, such as roughness measurements, for example, using the classical Bennett’s formula. Measurements were performed using two experimental goniophotometers to fully characterize the BRDF of tungsten samples with different roughness values. Surface topography was measured using a three-dimensional laser scanning confocal microscope. Several parameters were extracted from these measurements including the arithmetic average roughness (Ra), the root mean square roughness (RMS), the Surface Inclination Angle Distribution and furthermore its mean value δm and the power spectral density (PSD). The correlations of BRDF model parameters deduced from the measurements are compared with the previous topographic parameters. The initial results on several tungsten samples show that Ra, which is the usual measure of surface roughness, is not the most suitable metric to link with the reflection behavior of the surface. In contrast, the PSD and the surface inclination angle are interesting metrics for describing the reflected light.

A simple approach to estimate lake surface water temperatures in Polish lowland lakes

Study regionLowland lakes located at Poland. Study focusThere are thousands of lowland lakes in Poland, however, many of them don’t have observed data of water temperatures. How can we simply estimate lake surface water temperature (LSWT) is challenging, and a few studies tried to deal with this issue. In this study, we proposed a simple method using the air2water model for regional estimation of LSWT for Polish lowland lakes: Firstly, a literature review was conducted to collect parameter values of the air2water model applications in different lakes with diverse lake morphologies (e.g., mean depth). Secondly, correlations of model parameters with mean lake depth were analyzed. Finally, model parameters of the Polish lowland lakes (25 lakes with three lakes for testing) were derived using the corresponding mean lake depth and model performance was evaluated afterwards. New hydrological insights for the regionThe results showed that the proposed method performs generally well for the 25 Polish lowland lakes, better than the previous method. Reasonable results were obtained for three Polish lakes that were not used for regression analysis, what encourages application of the proposed approach on unmonitored lakes. The results indicated that the method proposed in this study can be used to estimate LSWT of Polish lowland lakes as daily data of air temperatures are readily available in most of the meteorological stations.

Stochastic simulation of velocity pulses of near-fault ground motions based on multivariate copula modeling

Due to the featured high-amplitude, long-period pulses in velocity or displacement time–history, the near-fault ground motion may have a stronger destructive effect on engineering structures. Therefore, studying velocity pulses modeling of near-fault ground motions has been an important topic in the earthquake engineering community. However, the variability and probabilistic correlation of pulse model parameters have not received enough attention. To this end, an elegant method for stochastic simulation of velocity pulses of near-fault ground motions is proposed based on the classical Gabor wavelet modeling and probabilistic correlation analysis of associated model parameters by multivariate copula modeling. Especially, an efficient scheme of iteratively determining the optimal copulas pertinent to edges of a complex vine structure is developed, and two parameters relevant to the near-fault ground motion records, i.e., the rupture distance and the strongest pulse orientation, are introduced which provides a physical basis for addressing the dependence structure among variables. For illustrative purposes, the ground accelerograms recorded by near-fault instruments in the 1999 Chi-Chi earthquake are used. It is shown that the multivariate copula modeling enables the efficient representation of probabilistic correlation between model parameters which is consistent with their physical mechanism. The time histories and spectral characteristics of velocity pulses of near-fault ground motions and their stochastic fluctuations with respect to the rupture distance are well simulated. While the empirical linear regression model does not meet the requirements of random vibration analysis and thus cannot guide the safety design of seismic structures.

Modeling gravity filling of dies on a rotary tablet press

For tableting on rotary presses, die filling is the quality dominating process step as the tablet weight and its variation is influenced. Therefore, a deeper understanding of this process is necessary. In this study, the filled powder weight was investigated for various parameter combinations on a rotary tablet press in lab-scale. Based on the obtained data for four materials, which differ in density and flow properties, a process-property model was developed to describe the volume flow into the dies. The flow can be transferred either into the final tablet weight or into the minimal paddle speed to fill the dies completely at a set turret speed, which supports process development. Finally, model parameters were correlated with the materials bulk density and permeability. Therefore, this models enables the prediction of the die filling process only based on geometrical data of the rotary press and measureable material properties, for the first time. • Investigation of the gravity filling process on a rotary tablet press in lab-scale. • Developed model to describe gravity filling process. • Correlation of model parameters with powder properties like permeability and bulk density. • Prediction of the minimal paddle speed to fill dies completely for a set turret speed by using the model. • Very good agreement between predicted and experimentally determined minimal paddle speed for all investigated materials.

Generative compressed breast shape model for digital mammography and digital breast tomosynthesis.

Modelling of the 3D breast shape under compression is of interest when optimizing image processing and reconstruction algorithms for mammography and digital breast tomosynthesis (DBT). Since these imaging techniques require the mechanical compression of the breast to obtain appropriate image quality, many such algorithms make use of breast-like phantoms. However, if phantoms do not have a realistic breast shape, this can impact the validity of such algorithms. To develop a point distribution model of the breast shape obtained through principal component analysis (PCA) of structured light (SL) scans from patient compressed breasts. SL scans were acquired at our institution during routine craniocaudal-view DBT imaging of 236 patients, creating a dataset containing DBT and SL scans with matching information. Thereafter, the SL scans were cleaned, merged, simplified, and set to a regular grid across all cases. A comparison between the initial SL scans after cleaning and the gridded SL scans was performed to determine the absolute difference between them. The scans with points in a regular grid were then used for PCA. Additionally, the correspondence between SL scans and DBT scans was assessed by comparing features such as the chest-to-nipple distance (CND), the projected breast area (PBA) and the length along the chest-wall (LCW). These features were compared using a paired t-test or the Wilcoxon signed rank sum test. Thereafter, the PCA shape prediction and SL scans were evaluated by calculating the mean absolute error to determine whether the model had adequately captured the information in the dataset. The coefficients obtained from the PCA could then parameterize a given breast shape as an offset from the sample means. We also explored correlations of the PCA breast shape model parameters with certain patient characteristics: age, glandular volume, glandular density by mass, total breast volume, compressed breast thickness, compression force, nipple location, and centre of the chest-wall. The median value across cases for the 90th and 99th percentiles of the interpolation error between the initial SL scans after cleaning and the gridded SL scans was 0.50 and 1.16mm, respectively. The comparison between SL and DBT scans resulted in small, but statistically significant, mean differences of 1.6mm, 1.6mm, and 2.2cm2 for the LCW, CND, and PBA, respectively. The final model achieved a median mean absolute error of 0.68mm compared to the scanned breast shapes and a perfect correlation between the first PCA coefficient and the patient breast compressed thickness, making it possible to use it to generate new model-based breast shapes with a specific breast thickness. There is a good agreement between the breast shape coverage obtained with SL scans used to construct our model and the DBT projection images, and we could therefore create a generative model based on this data that is available for download on Github.

An ANN-based generic energy model of cleanroom air-conditioning systems for high-tech fabrication location and technology assessments

The air-conditioning systems for high-tech clean fabrications are usually energy-intensive as they require strict space temperature, humidity and cleanliness controls simultaneously. Given that their energy performance is significantly affected by climate conditions and application scenarios concerning production technologies used, simple, reliable and widely-applicable energy models for effective fabrication location and technology assessments are of great value for decision-makers. However, no such model can be found in previous studies. In this paper, an ANN-based generic energy model for full climates and full application scenarios is proposed for location and technology assessments of high-tech cleanrooms at planning stage. The model is developed based on energy data of a typical cleanroom under all climates in China and all application scenarios by detailed simulations. The correlation of model parameters with climates and application scenarios is identified using a regression method and an ANN method. The model is validated in 22 worldwide cities under 420 application scenarios, showing satisfactory accuracy. A comprehensive energy assessment study on high-tech cleanrooms in 31 major Chinese cities is conducted under typical scenarios, providing clear overall energy use pictures of locating clean fabs. Results show that proper location selection could provide potential energy savings of 30% for large-scale clean fabs.

Variation in thermal and hydrotime requirements for seed germination of Chaco seasonally dry forest species in relation to population environmental conditions and seed mass

AbstractThermal and hydrotime requirements for seed germination can vary with the species, environment in which the seeds developed and/or seed mass. We investigated seed germination responses to temperature and water potential of Aspidosperma quebracho‐blanco, Bulnesia retama, Parasenegalia visco, Parkinsonia praecox and Tecoma stans, all of which are native tree species of Chaco seasonally dry forests in northwest Argentina. We determined inter‐and intraspecific variability in seed germination through thermal and hydrotime models and correlated model parameters with environmental variables of seed provenance and seed mass. Seeds of all species germinated to high percentage over a range of temperatures. Seeds of A. quebracho‐blanco, P. praecox and T. stans were more tolerant to water stress than those of B. retama and P. visco. Thermal and hydrotime parameters for germination varied amongst species and populations. There was a negative correlation between thermal (θT(50)) and hydrotime (θH) constants and base temperature (Tb) and water potential (Ψb(50)). Seed mass correlated negatively with Tb and positively with θT(50). Species and populations were grouped according to thermal and hydrotime parameters and seed mass but not according to seed provenance. These results suggest that seed germination traits related to temperature, water potential and seed mass are species specific, and the variations within a species are probably due to a high phenotypic plasticity. These species showed a complex set of germination strategies that give them a competitive advantage in different environments, depending on dispersal time, seed mass and local environmental conditions during the germination period.

Apparent diffusion coefficient and intravoxel incoherent motion-diffusion kurtosis model parameters in invasive breast cancer: Correlation with the histological parameters of whole-slide imaging

ObjectivesTo evaluate the correlation of ADC and IVIM-diffusion kurtosis (DK) model parameters with quantitative histological parameters of whole-slide imaging (WSI). MethodsThis retrospective study (September 2015–July 2016) included 37 consecutive patients (all females; median age: 64 years, range, 41–82 years), each with a single invasive breast ductal carcinoma lesion with a mass appearance on preoperative MRI. DWI with b-values of 0, 50, 100, 300, 550, 850, and 1000 s/mm2 was performed. ADC maps were generated with b-values of 50 and 850 s/mm2. The IVIM-DK model was analysed using the following formula:Sb=S0fe−bD∗+1−fe−bD+1/6kbD2where S is the signal intensity, b is the b-value, f is the perfusion fraction, D* is the pseudo-diffusion coefficient of the vascular component, D is the diffusion coefficient of the non-vascular component, and k is the diffusion kurtosis. Whole tumour segmentation was performed to obtain the mean ADC, f, D*, D, and k. Quantitative histological parameters were obtained using cytokeratin immunostaining and WSI. The correlation of ADC and IVIM-DK model parameters with quantitative histological parameters was examined by Pearson's correlation. ResultsThe ADC was significantly correlated with the area ratio of interstitium (r = 0.53, p = 0.00082) and entropy (r = −0.58, p = 0.00019). k was significantly correlated with the area ratio of cancer cell nuclei (r = 0.53, p = 0.00079). ConclusionsSince the ADC reflected the area ratio of interstitium and entropy, and diffusion kurtosis reflected the area ratio of cancer cell nuclei, these parameters may be effective in distinguishing between benign and malignant breast tumours and in grading breast cancer.

Work hardening in metastable high entropy alloys: a modified five-parameter model

Excellent work hardening in transformation-induced plasticity (TRIP)-enabled metastable high entropy alloys (HEAs) owe to persistent austenite (γ) to martensite (ε) phase transformation; non-basal slip activity and deformation twinning in the transformed martensitic phase are additional deformation mechanisms that contribute to work hardening in selected TRIP HEAs. Mechanical response of TRIP HEAs under uniaxial tension is characterized by an intermediate stage distinguished by a gradual increase in work hardening rate preceded and succeeded by stages of rapid drop in the work hardening rate. A five-parameter empirical model that replicates the nature of the work hardening rate curve in TRIP HEAs has been developed. The ease of parametric identification of the model from the global stress-strain response simplifies implementation of the model over physically based models. As the propensity of stress induced transformation in TRIP HEAs is related to the stacking fault energy (SFE) of the parent austenite phase, an attempt is made to correlate the model parameters with SFE. Based on the trends indicated in the correlation of model parameters with SFE as well as the initial microstructure, a method is proposed to predict the tensile stress-strain response of TRIP HEAs.

Correlations of Rock-Physics Model Parameters From Bayesian Analysis: Pressure- and Porosity-Dependent Models Applied to Unconsolidated Sands

Correlations of rock-physics model inputs are important to know to help design informative prior models within integrated reservoir-characterization workflows. A Bayesian framework is optimal to determine such correlations. Within that framework, we use velocity and porosity measurements on unconsolidated, dry, and clean sands. Three pressure- and three porosity-dependent rock-physics models are applied to the data to examine relationships among the inputs. As with any Bayesian formulation, we define a prior model and calculate the likelihood in order to evaluate the posterior. With relatively few inputs to consider for each rock-physics model, we found that sampling the posterior exhaustively to be convenient. The results of the Bayesian analyses are multivariate histograms that indicate most likely values of the input parameters given the data to which the rock-physics model was fit. When the Bayesian procedure is repeated many times for the same data, but with different prior models, correlations emerged among the input parameters in a rock-physics model. These correlations were not known previously. Implications, for the pressure- and porosity-dependent models examined here, are that these correlations should be utilized when applying these models to other relevant data sets. Furthermore, additional rock-physics models should be examined similarly to determine any potential correlations in their inputs. These correlations can then be taken advantage of in forward and inverse problems posed in reservoir characterization.

Estimation of porosity, fluid bulk modulus, and stiff-pore volume fraction using a multitrace Bayesian amplitude-variation-with-offset petrophysics inversion in multiporosity reservoirs

The estimation of petrophysical and fluid-filling properties of subsurface reservoirs from seismic data is a crucial component of reservoir characterization. Seismic amplitude-variation-with-offset (AVO) inversion driven by rock physics is an effective approach to characterize reservoir properties. In general, PP-wave reflection coefficients, elastic moduli, and petrophysical parameters are nonlinearly coupled, especially in multiple-type pore-space reservoirs, which makes seismic AVO petrophysics inversion ill-posed. We have developed a new approach that combines Biot-Gassmann’s poroelasticity theory with Russell’s linear AVO approximation, to estimate the reservoir properties including elastic moduli and petrophysical parameters based on multitrace probabilistic AVO inversion algorithm. We first derive a novel PP-wave reflection coefficient formulation in terms of porosity, stiff-pore volume fraction, rock-matrix shear modulus, and fluid bulk modulus to incorporate the effect of pore structures on elastic moduli by considering the soft and stiff pores with different aspect ratios in sandstone reservoirs. Through the analysis of the four types of PP-wave reflection coefficients, the approximation accuracy and inversion feasibility of the derived formulation are verified. Our stochastic inversion method aims to predict the posterior probability density function in a Bayesian setting according to a prior Laplace distribution with vertical correlation and prior Gaussian distribution with lateral correlation of model parameters. A Metropolis-Hastings stochastic sampling algorithm with multiple Markov chains is developed to simulate the posterior models of porosity, stiff-pore volume fraction, rock-matrix shear modulus, and fluid bulk modulus from seismic AVO gathers. The applicability and validity of our inversion method is illustrated with synthetic examples and a real data application.

An analytic hybrid halo + perturbation theory model for small-scale correlators: baryons, halos, and galaxies

We update Halo Zeldovich Perturbation Theory (HZPT, [1]), an analytic model for the two-point statistics of dark matter, to describe halo and galaxy clustering, and galaxy-matter cross-correlation on nonlinear scales.The model correcting Zeldovich has an analytic Fourier transform, and therefore is valid in both configuration space and Fourier space.The model is accurate at the 2%-level or less for P_mm (k< 1 h/Mpc), P_hm (k< 1 h/Mpc), P_hh (k< 2 h/Mpc), Pgm (k< 1 h/Mpc), Pgg (k< 1 h/Mpc), ξmm (r> 1 Mpc/h), ξhm (r> 2 Mpc/h), ξhh (r> 2 Mpc/h), ξgm (r> 1 Mpc/h), ξgg (r> 2 Mpc/h), for LRG-like mock galaxies.We show that the HZPT model for matter correlators can account for the effects of a wide range of baryonic feedback models and provide two extended dark matter models which are of 1% (3%) accuracy for k < 10 (8) h/Mpc.We explicitly model the non-perturbative features of halo exclusion for the halo-halo and galaxy-galaxy correlators, as well as the presence of satellites for galaxy-matter and galaxy-galaxy correlation functions.We perform density estimation using N-body simulations and a wide range of HOD galaxy mocks to obtain correlations of model parameters with the cosmological parameters Ωm and σ8.HZPT can provide a fast, interpretable, and analytic model for combined-probe analyses of redshift surveys using scales well into the non-linear regime.

The influence of coupled kinetic-equilibrium in breaking the correlation between kinetic parameters

The most important requirement in nonlinear model fitting is an approach to solve the problem concerning the determination of full correlated model parameters. Since nonlinear model-based analysis is used in the chemical process, the importance of this issue is further increased. For instance, it is challenging to uniquely measure the rate constants in first-order reverse reactions from incomplete concentration data. This is due to the correlation between these constants. Therefore, breaking this correlation is necessary to determine the rate constants. In this work, the coupled kinetic-equilibrium process is suggested as a new strategy to overcome this problem. Also, this fact allows for determining kinetic parameters in multi-step kinetic reactions. The model-based global analysis is a complete analysis of the complex reaction mechanisms with the possibility of calculating and inserting the equilibrium and kinetic parameters into the fitting algorithm. After the fitting process, the optimal equilibrium and kinetic parameters are obtained, along with an estimate of their standard deviations. Therefore, simulated data sets for a range of different chemical cases, as well as the pH-dependent keto-enol tautomerization are used to exemplify the methods.

Methods of Identifying Correlated Model Parameters with Noise in Prognostics

In physics-based prognostics, model parameters are estimated by minimizing the error or maximizing the likelihood between model predictions and measured data. When multiple model parameters are strongly correlated, it is challenging to identify individual parameters by measuring degradation data, especially when the data have noise. This paper first presents various correlations that occur during the process of model parameter estimation and then introduces two methods of identifying the accurate values of individual parameters when they are strongly correlated. The first method can be applied when the correlation relationship evolves as damage grows, while the second method can be applied when the operating (loading) conditions change. Starting from manufactured data using the true parameters, the accuracy of identified parameters is compared with various levels of noise. It turned out that the proposed method can identify the accurate values of model parameters even with a relatively large level of noise. In terms of the marginal distribution, the standard deviation of a model parameter is reduced from 0.125 to 0.03 when different damage states are used. When the loading conditions change, the uncertainty is reduced from 0.3 to 0.05. Both are considered as a significant improvement.

Exploring the variance contributions of correlated model parameters: A sampling-based approach and its application in traffic simulation models

Analyzing the impacts of model parameters on model outputs is an important but challenging topic for scientific research involving simulation models. Global Sensitivity Analysis (SA) has been recently employed by many transportation researchers for such task, but a proper SA is still not a common practice. In particular, many modelers simply assume that all parameters are uncorrelated in the SA. However, this assumption is often unrealistic for traffic simulation models, in which many parameters are actually correlated, leading to wrong conclusions.In this paper, a sampling-based approach is provided for the SA of correlated parameters. It uses Gaussian copula to link the marginal distributions of individual parameters with their global distributions and correlations, and utilizes the extended Sobol’ formula to estimate the variance-based sensitivity indexes in a Monte Carlo framework.Its application is illustrated using two different car-following models: the Intelligent Driver Model (IDM) and the Wiedemann-74 (W74) model. Results show that this method is able to accurately quantify the sensitivity of all model parameters. As a general method, this approach can be transferred like a standard quantitative SA tool to any traffic model or complex model in the wider scientific community, especially when correlated parameters exist.

Ultra-fast Model Emulation with PRISM: Analyzing the Meraxes Galaxy Formation Model

Abstract We demonstrate the potential of an emulator-based approach to analyzing galaxy formation models in the domain where constraining data is limited. We have applied the open-source Python package Prism to the galaxy formation model Meraxes. Meraxes is a semianalytic model, purposely built to study the growth of galaxies during the Epoch of Reionization. Constraining such models is however complicated by the scarcity of observational data in the EoR. Prism’s ability to rapidly construct accurate approximations of complex scientific models using minimal data is therefore key to performing this analysis well. This paper provides an overview of our analysis of Meraxes using measurements of galaxy stellar mass densities, luminosity functions, and color–magnitude relations. We demonstrate the power of using Prism instead of a full Bayesian analysis when dealing with highly correlated model parameters and a scarce set of observational data. Our results show that the various observational data sets constrain Meraxes differently and do not necessarily agree with each other, signifying the importance of using multiple observational data types when constraining such models. Furthermore, we show that Prism can detect when model parameters are too correlated or cannot be constrained effectively. We conclude that a mixture of different observational data types, even when they are scarce or inaccurate, is a priority for understanding galaxy formation and that emulation frameworks such as Prism can guide the selection of such data.

Correlation of radiomic features on dynamic contrast-enhanced magnetic resonance with microvessel density in hepatocellular carcinoma based on different models.

ObjectiveTo explore the correlations of radiomic features of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) with microvessel density (MVD) in patients with hepatocellular carcinoma (HCC), based on single-input and dual-input two-compartment extended Tofts (SITET and DITET) models.MethodsWe compared the quantitative parameters of SITET and DITET models for DCE-MRI in 30 patients with HCC using paired sample t-tests. The correlations of SITET and DITET model parameters with CD31-MVD and CD34-MVD were analyzed using Pearson’s correlation analysis. A diagnostic model of CD34-MVD was established and the diagnostic abilities of models for MVD were analyzed using receiver operating characteristic curve (ROC) analysis.ResultsThere were significant differences between the quantitative parameters in the two kinds of models. Compared with SITET, DITET parameters showed better correlations with CD31-MVD and CD34-MVD. The Ktrans and Ve radiomics features of the DITET model showed high efficiency for predicting the level of CD34-MVD according to ROC analysis, with areas under curves of 0.83 and 0.94, respectively.ConclusionCompared with SITET, the DITET model provides a better indication of the microcirculation of HCC and is thus more suitable for examining patients with HCC.

Convolutional neural network with group theory and random selection particle swarm optimizer for enhancing cancer image classification.

As an epitome of deep learning, convolutional neural network (CNN) has shown its advantages in solving many real-world problems. Successful CNN applications on medical prognosis and diagnosis have been achieved in recent years. Their common goal is to recognize the insights from the subtle details from medical images by building a suitable CNN model with maximum accuracy and minimum error. The CNN performance is extremely sensitive to the parameter tuning for any given network structure. To approach this concern, a novel self-tuning CNN model is proposed with a significant characteristic of having a metaheuristic-based optimizer. The most optimal set of parameters is often found via our proposed method, namely group theory and random selection-based particle swarm optimization (GTRS-PSO). The insights of symmetric essentials of model structure and parameter correlation are extracted, followed by the hierarchical partitioning of parameter space, and four operators on those partitions are designed for moving neighborhoods and formulating the swarm topology accordingly. The parameters are updated by a random selection strategy at each interval of partitions during the search process. Preliminary experiments over two radiology image datasets: breast cancer and lung cancer, are conducted for a comprehensive comparison of GTRS-PSO versus other optimization algorithms. The results show that CNN with GTRS-PSO optimizer can achieve the best performance for cancer image classifications, especially when there are symmetric components inside the data properties and model structures.