Single Variable Model Research Articles

Multifactorial assessment of Parkinson's disease course and outcomes using trajectory modeling in a multiethnic, multisite cohort - extension of the LONG-PD study.

The severity, progression, and outcomes of motor and non-motor symptoms in Parkinson's disease (PD) are quite variable. Following PD cohorts holds promise for identifying predictors of disease severity and progression. PD patients (N = 871) were enrolled at five sites. Enrollment occurred within 5 years of initial motor symptom onset. Disease progression was assessed annually for 2-to-10 years after onset. Group-based trajectory modeling was used to identify groups differing in disease progression. Models were developed for UPDRS-III scores, UPDRS-III tremor and bradykinesia-rigidity subscores, Hoehn & Yahr (H&Y) stage, Mini-Mental Status Exam (MMSE) scores, and UPDRS-III, H&Y and MMSE scores considered together. Predictors of trajectory-group membership were modeled simultaneously with the trajectories. Kaplan-Meier survival analysis evaluated survival free of PD outcomes. The best fitting models identified three groups. One showed a relatively benign, slowly progressing trajectory (Group 1), a second showed a moderate, intermediately progressing trajectory (Group 2), and a third showed a more severe, rapidly progressing trajectory (Group 3). Stable trajectory-group membership occurred relatively early in the disease course, 5 years after initial motor symptom. Predictors of intermediate and more severe trajectory-group membership varied across the single variable models and the multivariable model jointly considering UPDRS-III, H&Y and MMSE scores. In the multivariable model, membership in Group 2 (28.4% of patients), relative to Group 1 (50.5%), was associated with male sex, younger age-at-onset, fewer education-years, pesticide exposure, absence of reported head injury, and akinetic/rigid subtype at initial presentation. Membership in Group 3 (21.3%), relative to Group 1, was associated with older age-at-onset, fewer education-years, pesticide exposure, and the absence of a tremor-predominant subtype at initial presentation. Persistent freezing, persistent falls, and cognitive impairment occurred earliest and more frequently in Group 3, later and less frequently in Group 2, and latest and least frequently in Group 1. Furthermore, autonomic complications, dysphagia, and psychosis occurred more frequently in Groups 2 and 3 than in Group 1. Modeling disease course using multiple objective assessments over an extended follow-up duration identified groups that more accurately reflect differences in PD course, prognosis, and outcomes than assessing single parameters over shorter intervals.

Relation between Modis-based Aerosol Optical Depth and Particulate Matter in Kathmandu using Regression Model

Ambient fine Particulate Matters have been linked to various adverse health outcomes. Exposure to the high level of such particles would increase the risk of premature death, especially for people with weak immune systems, such as children and elder people. This research derives the relation between particulate matter and AOD from the Regression model on the seasonal (Pre-monsoon season (March 2020) and winter season (December 2019) basis of Kathmandu. Here two models have been developed one linear single-variable regression model and the other multivariable regression model. For the multivariable regression model, meteorological factors like Wind speed, Temperature, and Relative Humidity were adopted from the underground and the Planetary boundary layer height was simulated from WRF. Particulate matter (PM2.5) was adopted from the US Embassy air quality station and MODIS Level 2 AOD having 10 km resolution was analyzed for regression modeling. The linear single variable and linear multivariable regression model were developed seasonally one from December 1st to December 31st, 2019 (winter season) and the other from March 1st to March 31st, 2020 (Pre-monsoon season) using Python. The seasonal correlation coefficient of these two models was obtained. In both seasons, the multivariable linear regression model showed a good correlation between AOD and Particulate Matter R2 (Pre-monsoon) = 0.72657, R2 (winter) = 0.4687) compared to the single variable regression model having R2 (Pre-monsoon) = 0.45, R2 (winter) = 0.133). In both these regression models using the evaluated regression coefficients, two seasonal equations were derived from which Particulate Matter can be estimated.

Mechanism of single variable grey forecasting modelling: Integration of increment and growth rate

Grey forecasting models as a kind of time series model have been widely applied in different areas, which have achieved high-precision performance for different series with evolutionary characteristics. This paper aims to deconstruct its modelling mechanism from the perspective of data characteristics. First, classical increment and growth rate relationship is analysed and a unified representation of the mixed relationship with increment and growth rate is given. To establish the relationship between grey forecasting models and traditional models, the mechanism is analysed by deconstructing discrete structural components. Then, the series of evolutionary characteristics in main grey forecasting models are linked and summarized. The affine properties of parameter estimation and predicted value are analysed before simulation. Furthermore, through numerical simulations, the performance of grey forecasting models for different feature sequences and the difference between single variable models is demonstrated. Finally, results in the real case show that MAPEs of grey forecasting models have controlled under 4% for different sequences, which are suitable for mixed feature sequences.

Irrigated corn grain yield prediction in Florida using active sensors and plant height

Remote sensing is widely utilized in agriculture for estimating corn (Zea mays L.) grain yield (CGY). Few studies have determined if the Normalized Difference Vegetation Index (NDVI) and/or Soil Plant Analysis Development (SPAD) can estimate CGY in Florida. From April to August 2022, in Live Oak, Florida, a field-scale experiment was conducted in two sites with irrigated corn using a complete randomized block design with six nitrogen (N) rates and four replicates per site. This study aimed to estimate CGY using NDVI alone or in combination with SPAD, plant height (PH), and N rate. CGY response curve served as a comparison standard. Fifteen data subsets were selected, and stepwise selection multiple linear regression analysis was utilized to generate each reduced equation (Model). In addition, the relative significance of the predictor variables was evaluated. The strongest correlations with CGY were demonstrated by N rate (r = 0.93), PH103 (r = 0.91), NDVI39 (r = 0.81), and SPAD60 (r = 0.93). Models with multiple variables showed a better fit than single-variable models. Model 15 (variables until tasseling - 60 DAP) demonstrated comparable performance with 92.8% of variance explained and RMSE = 1,315.685 kg ha−1. Regardless of the model, the N rate has always contributed the most to CGY. Although Model 1 had the best overall performance, it may not be feasible for growers to utilize a model with multiple terms. Consequently, Model 15 could estimate CGY in Florida based on PH and NDVI at 60 and 32 DAP, respectively.

T Cell Subsets are Associated with Clinical Activity and Efficacy of 4.5g Intravenous Glucocorticoid for Moderate-To-Severe Thyroid-Associated Ophthalmopathy

ABSTRACT Background Intravenous glucocorticoid (IVGC) remains the main treatment for moderate-to-severe and active thyroid-associated ophthalmopathy (TAO). However, a substantial number (20–30%) of active moderate-to-severe TAO patients may not respond to IVGC. Some patients may have disease progression despite IVGC treatment or relapse after steroid withdrawal. Objectives To analyze risk factors for clinical activity and predictive factors for clinical outcomes of 4.5 g IVGC therapy in patients with moderate-to-severe TAO. Design and methods Our study was performed in two steps: step 1 involved 110 moderate-to-severe TAO patients and analyzed risk factors for TAO activity; step 2 involved 53 active moderate-to-severe TAO patients from step 1 who were treated with 4.5 g IVGC therapy and analyzed predictive factors for clinical outcomes of IVGC therapy. Multivariate logistic regression analysis was used to identify the independent predictors and establish the predictive model. Results Abnormal TRAb (OR = 4.717; P = 0.019) and the percentage of CD3+CD4+ T cell (OR = 1.092; P = 0.028) were independently associated with the activity of moderate-to-severe TAO patients. The pretreatment CAS-max in both eyes (OR = 7.221; P = 0.013) and the percentage of pretreatment CD3+T cell (OR = 0.718; P = 0.037) were independently associated with therapeutic efficacy. The pretreatment CAS-max in both eyes (OR = 156.53; P = 0.028) and the percentage of post-treatment CD3+T cell (OR = 0.554; P = 0.043) were independently associated with therapeutic efficacy. Besides, multivariable prediction models were established, which were better in the forecasting aspect than single-variable prediction models. Conclusions Based on the findings of this study, we should monitor the peripheral blood T cell subsets for TAO, which could be helpful to timely judge the condition of clinical manifestation and effect of treatment for TAO patients. Multivariable prediction models have been established, which have great significance for clinical work.

Periodic orbits of the perturbed relative motion

This paper presents a novel dynamical system for the perturbed relative motion of deputy satellite. The governing equations are derived in the LVLH frame under the second zonal harmonic coefficient effect of the Earth on both chief and deputy satellites. An extended numerical analysis on the deputy satellite motion is performed, where the initial state vector of deputy satellite and eccentricity of chief satellite are used as initial conditions for finding its own periodic orbits. The interior loops of periodic orbits are obtained using numerical integrator RKEOM. Such orbits are analysed for different parameters and their stability are also discussed, where these orbits have considerable importance in the resonance analysis of motion The second zonal harmonic coefficient effect on the periodicity, initial position, initial true anomaly is studied, alongside exploring these orbits with one, two and three loops. Furthermore, a single variable nonlinear regression model is developed to predict initial true anomaly using eccentricity of chief satellite’s orbit.

Association Between Fuchs Endothelial Corneal Dystrophy, Diabetes Mellitus, and Multimorbidity.

The aim of this study was to assess risk for demographic variables and other health conditions that are associated with Fuchs endothelial corneal dystrophy (FECD). We developed a FECD case-control algorithm based on structured electronic health record data and confirmed accuracy by individual review of charts at 3 Veterans Affairs (VA) Medical Centers. This algorithm was applied to the Department of VA Million Veteran Program cohort from whom sex, genetic ancestry, comorbidities, diagnostic phecodes, and laboratory values were extracted. Single-variable and multiple variable logistic regression models were used to determine the association of these risk factors with FECD diagnosis. Being a FECD case was associated with female sex, European genetic ancestry, and a greater number of comorbidities. Of 1417 diagnostic phecodes evaluated, 213 had a significant association with FECD, falling in both ocular and nonocular conditions, including diabetes mellitus (DM). Five of 69 laboratory values were associated with FECD, with the direction of change for 4 being consistent with DM. Insulin dependency and type 1 DM raised risk to a greater degree than type 2 DM, like other microvascular diabetic complications. Female sex, European ancestry, and multimorbidity increased FECD risk. Endocrine/metabolic clinic encounter codes and altered patterns of laboratory values support DM increasing FECD risk. Our results evoke a threshold model in which the FECD phenotype is intensified by DM and potentially other health conditions that alter corneal physiology. Further studies to better understand the relationship between FECD and DM are indicated and may help identify opportunities for slowing FECD progression.

Modelling the soil C impacts of cover crops in temperate regions

CONTEXTAgricultural land management decisions are based on numerous considerations. Belowground carbon (C) storage for both ecosystem health and greenhouse gas (GHG) management is a growing motivation. Observed heterogeneity in soil C storage in croplands may be driven by various environmental, climatic and management factors. Farm system models can indicate which practices will drive C storage, provided the practice is well parameterised and the land manager can provide necessary input data. OBJECTIVEWe aimed to predict soil C impacts of temperate cover cropping using simple models suitable for broad farmer use and decision support. METHODSThe dataset used was initially compiled for a meta-analysis (McClelland et al., 2021) to quantify soil C response to cover crop treatments relative to a non-cover cropped system. It contains 181 data points from 40 existing studies in temperate climates. Environmental, climatic and management indicators were regressed pairwise to predict annual soil C stock change under cover cropping relative to no cover cropping. We also included the IPCC tier 1 methodology and meta-analysis response ratios in our model comparison.The ease of reliable measurement and monitoring across the modelled indicators was also considered because the best-correlated relationships are squandered if data constraints risk decision-makers being unable to use the model. RESULTS AND CONCLUSIONSUsing an extended test dataset to consider priorities for model users, several regression models outperformed the IPCC tier 1 methodology. In particular, two regression models reliably predicted negative changes in soil C, which IPCC and meta-analysis factor approaches could not. A single variable regression model based on cover crop biomass (dry matter) production was the best combination of statistical power, biological relevance and parsimony. In temperate climates, we predicted an increase in soil C stocks as long as cover crop biomass production exceeded 1.3 Mg ha−1 yr−1. SIGNIFICANCEOur final model can be applied with estimated user input data, and avoids the need for baseline soil C as an input; this makes it relatively accessible for farmers. Parsimonious models for soil C change under land management practices can be effective and are an opportunity to increase access to soil C management information for farmers.

Atomic structure evolution and linear regression fitting models for pre-breakdown electric field strength of FCC, BCC and HCP metal nano-emitters under high electric field from PIC-ED–MD simulations

Multi-scale and multi-physics simulations are carried out for nano-emitters consisting of FCC (Al, Cu and Au), BCC (V, Mo and W) and HCP (Ti, Zn and Zr) metals, using hybrid electrodynamics coupled with molecular dynamics-particle in cell simulations (PIC-ED–MD). We show that the tilting of the nano-emitter at low temperature and small electric field (E-field) is mainly caused either by the movement of partial dislocations at the apex of the nanotip or by the elastic local distortions of atomic registries away from their ideal lattice sites (FCC/BCC/HCP). At high E-field, the intense resistive heating due to the strong electron emission leads to the direct melting of the apex of nano-emitters. For nano-emitters consisting of low melting point metals such as Al, Zn and Au, the thermal runaway is driven by the elongation, thinning and necking of the molten region. Meanwhile, the elongation, thinning and sharpening produce the nano-protrusion at the apex of metal nano-emitters, and the detachment of atoms or atomic clusters from the nano-protrusion mainly contributes to the thermal runaway event for refractor metals such as Ti, Zr, Mo and W. The critical E-field strength of metal nano-emitters is found to be strongly correlated with structural parameters (atomic coordination number of liquid and equilibrium lattice constant), thermodynamic quantities (cohesive energy and enthalpy of evaporation) and phase transition temperatures (melting point and boiling point). These correlations enable us to establish either single-variable linear fitting models or multi-variable linear regression models to predict the critical E-field value for metal nano-emitters with good credibility.

Using multiple linear regression for estimating total mussel (Mytilus californianus) shell size to accommodate morphological variability related to diverse environments

Understanding variation in mollusk shell size through time can be useful for interpreting changes in past marine environments and human behavior related to subsistence. Recent studies have provided means to estimate total shell length for California mussel (Mytilus californianus) by using a single morphological measurement in a simple linear regression approach; however, shell morphology may differ between environmentally diverse areas and through time, in turn potentially influencing the reliability of single variable models generated from measurements of shells harvested from a different geographic region to predict total shell length for archaeological samples. In this study, we examined the performance of models built from three site specific samples from contemporary mussel beds from Santa Rosa Island, southern California and models built from a pooled sample. The data was split into training (80%) and testing (20%) subsets to construct and validate eight multivariable models using a best subset regression approach. We then examined the performance of our models on each subset, evaluating measures of bias, accuracy, coverage, and precision. We also evaluated the performance of our models using an archaeological sample (CA-SRI-138) as an additional external validation sample. For each sample, we consider a model with all measurements and a model using only those measurements likely observed in fragmentary assemblages. We also developed single variable models for comparison to the performance of multivariable models. Multivariable models consistently demonstrate better performance than single variable models. The models using the pooled sample were also more accurate, precise, and exhibited less bias than the site-specific models. Each site-specific model selected a different subset of variables and indicated a range of variable importance, which is captured in the increased generalizability of the multivariable, pooled sample models when estimating total shell length for both contemporary and archaeological samples.

Abstract 1973: A geo inelegant approach to cancer disparity

Abstract UCRShiny is an intelligent user interface application that utilizes R and R-Shiny tools and statistical intelligence to allow users to visualize geospatial cancer data and perform advanced analyses and predictive modeling. UCRShiny's focus is the catchment area (in Illinois and Indiana) near UCCC and utilizes various metadata. This includes fitting univariate and multivariate models to make statistically sound inferences from the data. It also allows the user to plot specific data on a map of the catchment area as layers for descriptive analysis and visual geospatial assessment. As well as assessing the significance of the relation between these factors. These statistical tools should have a level of intelligence that can determine the correctness of the model's assumptions. An advantage of our tool is that it provides an overall diagnosis of the model to the user without the need for the intervention of an expert statistician. The data selected for analysis are assessed to ensure they meet the assumptions of the statistical model being used, and proper data transformations are applied. The presence of confounding factors is assessed using linear models in sequence. Initially, a single variable linear model is fit using the primary variable of interest (i.e., Cancer mortality) as the response variable and the second variable in the univariate analysis as the covariate. The second model adds the internal metadata columns as potential confounders to the model. In the third step, the percent change in the coefficient of the second variable in the first and second linear models is calculated. Multivariate analysis using ML is used to simultaneously assess the effect of multiple variables on the outcome variable. Here, we use various ML algorithms based on the problem and data at hand, report the significant features, and quantify their impact on the response of interest. The tool allows users to utilize the in-app data, load their dataset, or combine both datasets. It helps to visualize geospatial health data on a map, perform statistical analysis and build ML/AI models without being concerned about the validity of the statistical models. After uploading data and selecting the appropriate statistical test/analysis, results will be visualized as a plot on a map with the specified layers collared by the layer variable values. The layers can be turned on or off using specific buttons on the map. GIS maps may be easily shared and included in apps and are available to almost everyone, anywhere. The user can study the effect of multiple variables and rank their importance in predicting the response variable using multivariate ML/AI models. The UI allows the user to select an appropriate ML algorithm for the model or the application's intelligent backend algorithms to choose the proper ML model (This will be implemented in phase II). The outcome feature and the covariates to use in the model can be selected using searchable textboxes in the UI, and the model is run using the Run model button. Citation Format: Ahmed Fadiel Metwaly, Adam Thomas Koster, Mohammad Abbasi, Aya Hassouneh, Jasmin Tiro, Nita Lee, Adekunle Odunsi. A geo inelegant approach to cancer disparity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1973.

Quantitative parameters in novel spectral computed tomography: Assessment of Ki-67 expression in patients with gastric adenocarcinoma.

The level of Ki-67 expression has served as a prognostic factor in gastric cancer. The quantitative parameters based on the novel dual-layer spectral detector computed tomography (DLSDCT) in discriminating the Ki-67 expression status are unclear. To investigate the diagnostic ability of DLSDCT-derived parameters for Ki-67 expression status in gastric carcinoma (GC). Dual-phase enhanced abdominal DLSDCT was performed preoperatively in 108 patients with gastric adenocarcinoma. Primary tumor monoenergetic CT attenuation value at 40-100 kilo electron volt (kev), the slope of the spectral curve (λHU), iodine concentration (IC), normalized IC (nIC), effective atomic number (Zeff) and normalized Zeff (nZeff) in the arterial phase (AP) and venous phase (VP) were retrospectively compared between patients with low and high Ki-67 expression in gastric adenocarcinoma. Spearman's correlation coefficient was used to analyze the association between the above parameters and Ki-67 expression status. Receiver operating characteristic (ROC) curve analysis was performed to compare the diagnostic efficacy of the statistically significant parameters between two groups. Thirty-seven and 71 patients were classified as having low and high Ki-67 expression, respectively. CT40 kev-VP, CT70 kev-VP, CT100 kev-VP, and Zeff-related parameters were significantly higher, but IC-related parameters were lower in the group with low Ki-67 expression status than the group with high Ki-67 expression status, and other analyzed parameters showed no statistical difference between the two groups. Spearman's correlation analysis showed that CT40 kev-VP, CT70 kev-VP, CT100 kev-VP, Zeff, and nZeff exhibited a negative correlation with Ki-67 status, whereas IC and nIC had positive correlation with Ki-67 status. The ROC analysis demonstrated that the multi-variable model of spectral parameters performed well in identifying the Ki-67 status [area under the curve (AUC) = 0.967; sensitivity 95.77%; specificity 91.89%)]. Nevertheless, the differentiating capabilities of single-variable model were moderate (AUC value 0.630 - 0.835). In addition, the nZeff VP and nICVP (AUC 0.835 and 0.805) showed better performance than CT40 kev-VP, CT70 kev-VP and CT100 kev-VP (AUC 0.630, 0.631 and 0.662) in discriminating the Ki-67 status. Quantitative spectral parameters are feasible to distinguish low and high Ki-67 expression in gastric adenocarcinoma. Zeff and IC may be useful parameters for evaluating the Ki-67 expression.

Model Updating for a Continuous Concrete Girder Bridge Using Data from Construction Monitoring

Finite element (FE) model updating can guarantee the accuracy of the numerical analyses for civil structures. However, two deficiencies currently exist with the traditional FE model updating technique based on the measurements of modal parameters and/or the static structural responses of the built structure with respect to its reference information insufficiency and its non-unique solution generally obtained, hampering its extensive use. It becomes the goal of the whole engineering community to introduce new effective methods for the civil structural FE model update. To this end, an innovative FE model updating method using data from construction monitoring is proposed in this article. With regard to its theoretical novelty, the new method transforms the complicated multi-variable optimization mathematical problem with the traditional FE model updating technique into many simple single-variable parameter identification problems. Under the engineering background of Huangsha Harbor Bridge, a three-span concrete continuous box girder bridge constructed utilizing the symmetric cantilever casting method, the effectiveness and the efficiency of the new model updating practice were validated. It is demonstrated using quantitative data that the abundant data measured on Huangsha Harbor Bridge in construction stages can enhance the reference information for the more accurate FE model updating of the structure, and the uncertain parameters with the initial FE model of Huangsha Harbor Bridge can be progressively and easily identified for the proposed model updating method using many single-variable linear regression models, instead of one complicated multi-variable mathematical or numerical model employed by the traditional model updating approaches, which generally leads to non-unique solutions rendered by normal optimization algorithms.

Comparison of constant and variable discharge flow and force coefficients for reciprocating compressor

Enhancing the mathematical models always have been a critical issue in every field, and the reciprocating compressors are not exceptional. Flow and force coefficients are two empirical factors that help to calculate the mass flow rate through the valve and gas force, respectively. These coefficients can be considered constant values or valve lift dependent. However, in the discharge process, the piston is close to the top dead center, which may affect the discharge flow field and, subsequently, the coefficients.These coefficients should be determined by experiments or numerical simulations such as computational fluid dynamics (CFD). Experiments are expensive, and the exact geometry may not always be available to perform the CFD analysis. Also, CFD analysis for various valve lifts and piston positions is computationally expensive.This study studied the effect of constant, single variable (valve lift dependent) and double variable (valve lift and piston position dependent) coefficients on compressor performance and discharge valve movement in three operating conditions. A mathematical model was developed for this purpose, and a 3-D CFD simulation was used to compute the coefficients. To validate the results, a strain gauge and a calorimeter were used as measurement tools.The single and double variable models captured the valve behavior, and pressure picks better than the constant model. However, the constant model estimated the indicated power, cooling capacity, and COP with good accuracy. In predicting the discharge losses, the double variable model was the most reliable among the others.

Physicochemical characterization of elephant apple (Dillenia indica L.) fruit and its mass and volume modeling using computer vision

The elephant apple (Dillenia indica L.) is an underutilized fruit and a potential source of many nutrients. The variations in the mass and volume of fruits are highly dependent on their physical properties. This relation is used to design precise post-harvest machinery and analyze fruit quality. The work aims to study the physicochemical properties and predict the correlation of mass and volume with various physical properties measured by image processing techniques through single and multivariable regression models such as linear, quadratic, rational, and exponential, respectively. The total phenolic content, flavonoid, and antioxidant activity of elephant apple powder are 1.2 mg GAE/g, 353.58 mg QE/g, and 5.74 mg GAEAC/g, respectively. Among single-variable models, the length-based rational model was best suited for predicting the mass and volume of elephant apples with high R2 values of 0.930 and 0.924, respectively. For mass and volume of combined physical properties, rational and exponential models showed the best fit with the R2 values of 0.935 and 0.950 with RMSE values of 18.196 and 15.638. Thus, the correlation between the mass and volume of elephant apples with their physical characteristics resulted in a quicker, more accurate, and more precise grading system.

Allometric models for aboveground biomass estimation of small trees and shrubs in African savanna ecosystems

Quantification of plant biomass and carbon in ecosystems is critical for establishing climate change mitigation potential. For large trees in various ecosystems, allometric models for estimating biomass have been developed but few biomass equations exist for small trees and shrubby vegetation. Allometric above-ground biomass (AGB) models are needed for small trees and shrubs in order to improve the quantification of biomass, particularly for savanna ecosystems, where small trees and shrubs comprise a significant portion of the biomass. In this study we have developed species-specific and multi-species allometric models for biomass estimation of small tree species and shrubs in the savanna ecosystem of Lake Mburo National Park in South Western Uganda. For our models we selected 27 small tree species (N = 403 individuals) and 12 shrub species (N = 177) common in savanna ecosystems for destructive sampling. We developed species-specific and multi-species allometric AGB models to provide estimates of AGB using specific biometric variables recorded for the small trees (i.e. species, DBH, height and crown area), and shrubs (species, height and crown area). We found that crown area was the best single predictor of species-specific AGB for small trees and for species-specific and multi-species models for shrubs. Species-specific models had the best predictive capacity of AGB compared to multi-species biomass models for small trees and shrubs. Multiple-variable models had the best predictive capacity of AGB in both species-specific and multi-species modeling compared to single-variable models. Based on these findings we conclude that the evaluation of carbon stocks of tropical savanna ecosystems should use multi-variable species-specific models for AGB estimation at the individual level, and multi-species models for AGB at the ecosystem level.

Predictors for late genitourinary toxicity in men receiving radiotherapy for high-risk prostate cancer using planned and accumulated dose

Significant deviations between bladder dose planned (DP) and dose accumulated (DA) have been reported in patients receiving radiotherapy for prostate cancer. This study aimed to construct multivariate analysis (MVA) models to predict the risk of late genitourinary (GU) toxicity with clinical and DP or DA as dose-volume (DV) variables. Bladder DA obtained from 150 patients were compared with DP. MVA models were built from significant clinical and DV variables (p<0.05) at univariate analysis. Previously developed dose-based-region-of-interest (DB-ROI) metrics using expanded ring structures from the prostate were included. Goodness-of-fit test and calibration plots were generated to determine model performance. Internal validation was accomplished using Bootstrapping. Intermediate-high DA (V30-65 Gy and DB-ROI-20-50mm) for bladder increased compared to DP. However, at the very high dose region, DA (D0.003 cc, V75 Gy, and DB-ROI-5-10mm) were significantly lower. In MVA, single variable models were generated with odds ratio (OR)<1. DB-ROI-50mm was predictive of Grade≥1 GU toxicity for DA and DP (DA and DP; OR: 0.96, p: 0.04) and achieved an area under the receiver operating curve (AUC) of>0.6. Prostate volume (OR: 0.87, p: 0.01) was significant in predicting Grade 2 GU toxicity with a high AUC of 0.81. Higher DA (V30-65 Gy) received by the bladder were not translated to higher late GU toxicity. DB-ROIs demonstrated higher predictive power than standard DV metrics in associating Grade≥1 toxicity. Smaller prostate volumes have a minor protective effect on late Grade 2 GU toxicity.

Improvement of protective facilities in vinyl chloride units on liver injury status of occupational exposed group

To comprehensively evaluate the effectiveness of improvement of protective facilities in a vinyl chloride monomer(VCM) on promotion toward health status of occupational exposed group and safety production at poly-vinyl chloride(PVC) by comparing the liver function indicators and inspection result before and after the improvement, and to analyze the possible influential factors. The information collection concerning facilities improvement in 2013 and 2016, identification toward critical controlling points, data or detection result from occupational on-site survey and physical examination were originated in July 2020, and 227 VCM exposed workers and 179 others without chemical reagents exposure history from production factory with calcium carbide process in Tianjin City were selected as objects. The effectiveness toward improvement of protective facilities in critical controlling points that occupational exposed workers usually involve in were evaluated through comparison toward VCM concentration result, thiodiglycolic acid(TDGA) content in urine, indicators on liver function and biochemistry and liver ultrasound. At the mean time, both binary Logistic regression analysis was used to screen the possible factors that contributed to abnormal symptoms and single dependent variable general linear regression model was used to find out the mutual interaction among influential factors. Prior to improvement on protective facilities, the C_(TWA)of VCM exposed by all 8 positions in G had exceeded 1 to 2 folds of occupational exposure limit(OEL=10 mg/m~3) and the position of synthetic operator and field sampling man were ones exposed to 1, 2-DCE with the range from 50% to 1 OEL, others hazard factors were found to meet relative OELs. Next, the content of TDGA in urine of exposed group was found to correlate strongly to the average concentration of VCM(r=0.79, P&lt;0.05), and result of TDGA content, alanine aminotransferase(ALT), γ-glutamyl transpeptidase(GGT) and abnormal rate toward liver ultrasound test in exposed group were much higher than ones in control group with significant differences(P&lt;0.05 or P&lt;0.001). By contrast, after the improvement, the exposure level toward all identified hazard factors were declined to meet OEL levels with significant differences in TDGA content, and result of ALT, GGT and abnormal rates toward symptoms in liver ultrasound test such as liver calcification with thickened liver echo, peripheral fibrosis of the liver, multiple hepatic cysts were markedly lower than ones before improvement and still were higher than ones in control group(P&lt;0.05 or P&lt;0.001). Further, parameters of gender, length of employment, weekly contact time, VCM exposed level, TDGA content were all important contributing factors to abnormal symptoms in liver ultrasound test before and after improvement on protective facilities(P&lt;0.05 or P&lt;0.001), in which a significant interaction effect between gender and length of employment(F=4.028, P=0.044), weekly contact time and TDGA content(F=2.183, P=0.046) in urine were found in contribution analysis to abnormal symptoms in liver ultrasound test(P&lt;0.05). The improvement measure carried out in VCM facilities by this PVC production factories with calcium carbide process effectively reduced the ambient concentration of hazard factors mainly led by VCM, but workers might still be at risk of liver injury even under VCM exposure at relative lower level, which may referred to factors of long weekly contact time, long length of employment and high fat diet.

Development, Validation and Comparison of Artificial Neural Network and Logistic Regression Models Predicting Eosinophilic Chronic Rhinosinusitis With Nasal Polyps.

Chronic rhinosinusitis with nasal polyps (CRSwNP) can be classified into eosinophilic CRSwNP (eCRSwNP) and non-eosinophilic CRSwNP (non-eCRSwNP) by tissue biopsy, which is difficult to perform preoperatively. Clinical biomarkers have predictive value for the classification of CRSwNP. We aimed to evaluate the application of artificial neural network (ANN) modeling in distinguishing different endotypes of CRSwNP based on clinical biomarkers. Clinical parameters were collected from 109 CRSwNP patients, and their predictive ability was analyzed. ANN and logistic regression (LR) models were developed in the training group (72 patients) and further tested in the test group (37 patients). The output variable was the diagnosis of eCRSwNP, defined as tissue eosinophil count > 10 per high-power field. The receiver operating characteristics curve was used to assess model performance. A total of 15 clinical features from 60 healthy controls, 60 eCRSwNP and 49 non-eCRSwNP were selected as candidate predictors. Nasal nitric oxide levels, peripheral eosinophil absolute count, total immunoglobulin E, and ratio of bilateral computed tomography scores for the ethmoid sinus and maxillary sinus were identified as important features for modeling. Two ANN models based on 4 and 15 clinical features were developed to predict eCRSwNP, which showed better performance, with the area under the receiver operator characteristics significantly higher than those from the respective LR models (0.976 vs. 0.902, P = 0.048; 0.970 vs. 0.845, P = 0.011). All ANN models had better fits than single variable prediction models (all P < 0.05), and ANN model 1 had the best predictive performance among all models. Machine learning models assist clinicians in predicting endotypes of nasal polyps before invasive detection. The ANN model has the potential to predict eCRSwNP with high sensitivity and specificity, and is superior to the LR model. ANNs are valuable for optimizing personalized patient management.

Interpretable Long Short-Term Memory Networks for Crop Yield Estimation

Food security is at stake, with climate change heavily impacting agriculture and food production. In the present context of extreme events and changing conditions, developing advanced crop yield models can learn from all available information and provide interpretable predictions for decision-making is of paramount relevance. This work explores the potential and limitations of developing interpretable crop yield models using long short-term memory (LSTM) neural networks, which typically excel at extracting information from time series. LSTMs were designed and trained with multi-source satellite and meteorological time series over Continental US (CONUS) and corn, soybean, and wheat yield data from the US Department of Agriculture. Two recent attribution methods are used to interpret and extract knowledge from the developed models: integrated gradients (IG), based on back-propagation, and Shapley values (SHAP), based on perturbations. Our results show that (1) LSTM models achieved high accuracy (R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> > 0.56); (2) multi-source combinations outperformed single variable models in all crop models; (3) both attribution methods were equivalent in detecting essential drivers and their contribution; (4) satellite estimates of enhanced vegetation index (EVI) and vegetation optical depth (VOD) together with meteorological estimates of maximum temperature (TMX) were the most relevant input features for crop yield estimations; and finally, (5) we discovered critical periods of the crop growth cycle for the corn, soybean, and wheat models. The suggested strategy constitutes an important step towards modeling and understanding crop production systems and advancing in automatic data-driven and accountable field management.