Greater Predictive Ability Research Articles

Hydrothermal bio-oil yield and higher heating value of high moisture and lipid biomass: Machine Learning modeling and feature response behavior analysis

The yield and higher heating value (HHV) of bio-oil products are significant performance parameters for the hydrothermal conversion of high-water and high-lipid biomass. Machine learning (ML) modeling prediction is a fast and convenient means of obtaining performance parameters. An informative dataset with 243 samples was prepared, and two highly adapted ML algorithms were used: Random Forest (RF) and Extreme Gradient Boosting Tree (XGBoost). It is interesting to note that the developed ML models demonstrated great prediction ability; for example, the regression coefficient () of the XGBoost model for yield and HHV prediction was as high as 0.942 and 0.940, respectively. Furthermore, partial dependence plots and SHAP (SHapley Additive exPlanations) interpretability methodologies were adopted to address the main contributions of the feature identification and response behavior analysis of the features. The results demonstrated that the biomass composition had the greatest effect on bio-oil yield, with fat contributing up to 40%. In contrast, the elemental composition had the most significant effect on the HHV of bio-oil. Notably, hydrogen content affected the HHV of up to 4.5 units. The interaction response behavior showed that the interaction of the process parameters with feedstock properties was most common and significant. The information obtained from the response mechanism can be used to enhance the subsequent hydrothermal fuel preparation process for bio-oils.

A Machine Learning-Based Web Tool for the Severity Prediction of COVID-19.

Predictive tools provide a unique opportunity to explain the observed differences in outcome between patients of the COVID-19 pandemic. The aim of this study was to associate individual demographic and clinical characteristics with disease severity in COVID-19 patients and to highlight the importance of machine learning (ML) in disease prognosis. The study enrolled 344 unvaccinated patients with confirmed SARS-CoV-2 infection. Data collected by integrating questionnaires and medical records were imported into various classification machine learning algorithms, and the algorithm and the hyperparameters with the greatest predictive ability were selected for use in a disease outcome prediction web tool. Of 111 independent features, age, sex, hypertension, obesity, and cancer comorbidity were found to be associated with severe COVID-19. Our prognostic tool can contribute to a successful therapeutic approach via personalized treatment. Although at the present time vaccination is not considered mandatory, this algorithm could encourage vulnerable groups to be vaccinated.

Selecting potential biomarkers of plasma proteins in mares with endometritis.

Endometritis is a common condition in mares that causes significant economic loss. Lacking obvious clinical signs, the clinical diagnosis of endometritis in mares relies on case-by-case clinical examinations, which can be particularly inefficient in large-scale farms. Therefore, the identification of potential biomarkers can serve as a non-invasive and efficient screening technique for endometritis in mares. To compare the blood proteome between fertile mares and mares with endometritis to identify biomarkers potentially associated with the development of endometritis and validate their predictive potential. Observational and experimental study. Differentially expressed proteins were identified via Data Independent Acquisition (DIA) proteomic profiling in a screening cohort composed of eight healthy mares and eight mares with endometritis. Subsequently, enzyme-linked immunosorbent assay was employed that included a validation cohort of 40 healthy mares and 40 mares with endometritis to verify the accuracy and sensitivity of the identified proteins, thereby establishing a diagnostic threshold. In the screening cohort, 12 proteins were significantly differentially expressed between endometritis mares and healthy controls (p < 0.05, outside the 1/1.2 to 1.2-fold). In the validation experiment, all six screened proteins were assessed with area under the curve (AUC) >0.8. The samples displayed certain levels of individual heterogeneity, and the number of samples analysed was limited. Additionally, the identified biomarkers were primarily associated with generalised inflammation, which potentially limited their specificity for endometritis. Levels of plasma proteins are sensitive indicators of equine endometritis and potential tools for endometritis screening. In plasma, fetuin B, von Willebrand factor, vitamin K-dependent protein C, insulin-like growth factor binding protein 3, interleukin 1 receptor accessory protein, and type II cell cytoskeleton showed great predictive ability, with fetuin B being the best predictor (AUC = 0.93, 95% CI: 0.89-0.98), which performs better when combined with all six detected proteins (AUC = 1, 95% CI: 0.99-1.00).

The close association of micronutrients with COVID-19

ObjectivesThe present study was conducted to explore the performance of micronutrients in the prediction and prevention of coronavirus disease 2019 (COVID-19). MethodsThis is an observational case-control study. 149 normal controls (NCs) and 214 COVID-19 patients were included in this study. Fat-soluble and water-soluble vitamins were determined by liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis, and inorganic elements were detected by inductively coupled plasma-mass spectrometry (ICP-MS) analysis. A logistic regression model based on six micronutrients were constructed using DxAI platform. ResultsMany micronutrients were dysregulated in COVID-19 compared to normal control (NC). 25-Hydroxyvitamin D3 [25(OH)D3], magnesium (Mg), copper (Cu), calcium (Ca) and vitamin B6 (pyridoxic acid, PA) were significantly independent risk factors for COVID-19. The logistic regression model consisted of 25(OH)D3, Mg, Cu, Ca, vitamin B5 (VB5) and PA was developed, and displayed a strong discriminative capability to differentiate COVID-19 patients from NC individuals [area under the receiver operating characteristic curve (AUROC) = 0.901]. In addition, the model had great predictive ability in discriminating mild/normal COVID-19 patients from NC individuals (AUROC = 0.883). ConclusionsOur study showed that micronutrients were associated with COVID-19, and our logistic regression model based on six micronutrients has potential in clinical management of COVID-19, and will be useful for prediction of COVID-19 and screening of high-risk population.

Serum α-hydroxybutyrate dehydrogenase as a biomarker for predicting survival outcomes in patients with UTUC after radical nephroureterectomy.

We aimed to determine the prognostic value of α-hydroxybutyrate dehydrogenase (α-HBDH) in upper tract urothelial carcinoma (UTUC) patients after radical nephroureterectomy (RNU). We retrospectively enrolled the data of 544 UTUC patients at West China Hospital from May 2003 to June 2019. Cancer-specific survival (CSS) was the endpoint of interest. The optimal cutoff value of α-HBDH was identified by X-Tile program. After propensity score matching (PSM), we utilized Kaplan‒Meier curves to estimate survival and Cox proportional hazard model for risk assessment. A nomogram was built based on the results of multivariate analysis, and calibration curve, time-dependent receiver operating characteristic (ROC) curves and decision curve analysis were also performed to evaluate the predictive accuracy. Overall, 394 and 150 patients were divided into the α-HBDH-low group and α-HBDH -high group at the cutoff value of 158 U/L, respectively. After PSM, the two groups were well matched for all confounding factors. High α-HBDH was associated with inferior CSS (P = 0.006), and preoperative α-HBDH was an independent predictor for CSS (HR: 1.36; 95% CI:1.08, 1.80), especially in localized UTUC patients (HR: 2.04; 95% CI:1.11, 3.74). Furthermore, the nomogram based on α-HBDH achieved great predictive ability for CSS with areas under the curves of 0.800 and 0.778 for 3-year and 5-year CSS, respectively. Serum α-HBDH was a novel and reliable biomarker for predicting survival outcomes in UTUC patients after RNU but should be further explored.

Discussion of a method for analysing adolescent depression based on BiLSTM and Attentional Mechanisms

The incidence of depression in adolescents has been commonplace, but the number of people diagnosed with depression in hospitals is not large, so it is important to analyze people's textual content in normal times to determine whether they are in a depressive mood, and then intervene in a timely manner. The article firstly introduces the research results of previous scholars in sentiment analysis, and then briefly introduces the LSTM(Long Short-Term Memory) model, explains the operation formula of the model, and points out that BiLSTM(Bi-directional Long Short-Term Memory) has greater predictive ability than the LSTM model by taking the before and after contexts of the sentences into full consideration, and by extracting bi-directional semantic features. After that, the principle of the Attention Mechanism and the application of its derivative models, Self-Attention Mechanism, Multi-Attention Mechanism and Cross-Attention Mechanism, in different scenarios are analyzed, and finally, the advantages and disadvantages of BiLSTM and the attention mechanism are compared in the prediction of textual sentiment, and it is proposed that the BiLSTM-Attention model can play a stronger performance in the analysis of textual sentiment.

Who are optimal candidates for primary tumor resection in patients with metastatic gastric adenocarcinoma? A population-based study.

The research aimed to construct a novel predictive nomogram to identify specific metastatic gastric adenocarcinoma (mGAC) populations who could benefit from primary tumor resection (PTR). Patients with mGAC were included in the SEER database and divided into PTR and non-PTR groups. The Kaplan-Meier analysis, propensity score matching (PSM), least absolute shrink and selection operator (LASSO) regression, multivariable logistic regression, and multivariate Cox regression methods were then used. Finally, the prediction nomograms were built and tested. 3185 patients with mGAC were enrolled. Among the patients, 679 cases underwent PTR while the other 2506 patients didn't receive PTR. After PSM, the patients in the PTR group presented longer median overall survival (15.0 vs. 7.0 months, p < 0.001). Among the PTR group, 307 (72.9%) patients obtained longer overall survival than seven months (beneficial group). Then the LASSO logistic regression was performed, and gender, grade, T stage, N stage, pathology, and chemotherapy were included to construct the nomogram. In both the training and validation cohorts, the nomogram exhibited good discrimination (AUC: 0.761 and 0.753, respectively). Furthermore, the other nomogram was constructed to predict 3-, 6-, and 12-month cancer-specific survival based on the variables from the multivariate Cox analysis. The 3-, 6-, and 12-month AUC values were 0.794, 0.739, and 0.698 in the training cohort, and 0.805, 0.759, and 0.695 in the validation cohorts. The calibration curves demonstrated relatively good consistency between the predicted and observed probabilities of survival in two nomograms. The models' clinical utility was revealed through decision curve analysis. The benefit nomogram could guide surgeons in decision-making and selecting optimal candidates for PTR among mGAC patients. And the prognostic nomogram presented great prediction ability for these patients.

Epigenetic dysregulated long non-coding RNAs in renal cell carcinoma based on multi-omics data and their influence on target drugs sensibility.

Epigenetic modifications play a crucial role in cancer development, and our study utilized public data to analyze which leads to the discovery of significant epigenetic abnormalities in lncRNAs, offering valuable insights into prognosis and treatment strategies for renal carcinoma. Public data were obtained from the Cancer Genome Atlas (TCGA), International Cancer Genome Consortium (ICGC) and Gene Expression Omnibus (GEO) database. The analysis of the online public data was all completed in R software. We discovered a great number of epigenetic abnormalities of lncRNA in renal cancer, which is achieved by comparing the following modification and methylation of histone region changes on the promoter and enhancer of lncRNA: H3K27ac, H3K4me1, H3K4me3. As a result, 12 specific epigenetic disorders of lncRNA genes in renal cancer were identified. Finally, based on this lncRNA, we investigated the prognosis of renal cancer samples, among which 8 lncRNA can be seen as markers of prognosis in renal cancer, which had great prediction ability for ccRCC prognosis. Meanwhile, high risk score may pose response better to axitinib and nilotinib, but not sorafenib or sunitinib. Beyond, we observed an elevated level of risk score in immunotherapy non-responders. Further, biological enrichment and immuno-infiltration analysis was conducted to investigate the fundamental differences between patients categorized as high or low risk. Our research improves the understanding in the function of epigenetic dysregulated long non-coding RNAs in renal carcinoma.

Predicting minty compounds binary mixtures' pleasantness by odor intensity in aqueous solutions.

The aroma of mint is well-liked by the public, and key flavor odorants in mint aroma had been found, but how these molecules interact and form a satisfying odor remains a challenge. Quality, intensity, and pleasantness are our most basic perceptions of aromas; both intensity and pleasantness can be quantified. However, compared to intensity, research on pleasantness was lacking. Pleasantness was one of the most important indicators for formulating a satisfying mint flavor, and the study of binary mixtures was fundamental to our understanding of more complex mixtures. Therefore, the purpose of this study was to explore the characteristics of pleasantness as a function of concentration and, at the same time, to investigate the relationship between intensity and pleasantness in binary mixtures. Thirty sensory evaluation volunteers participated in the evaluation of the intensity and pleasantness of six key flavor odorants of mint and five binary mixtures. The results showed that the pleasantness increased first and then decreased or stabilized with the rising of concentration; even though the interactions in binary mixtures were not the same, their pleasantness could be predicted using the intensities of the components by Response Surface Design of Experiments, and the goodness of fit was greater than 0.92, indicating that the models had the great predictive ability. PRACTICAL APPLICATION: Whether blending flavors or evaluating them, a great deal of experience is required, yet the acquisition of this experience is a long process. Performing these tasks is difficult for the novice, and it helps to quantify the feeling for the flavor and build some mathematical models.

Forecast and Analysis of a Rainstorm Case in East China Based on the Blown-Up Theory

Practical application has shown that the blown-up theory has great predictive ability for predicting transitional weather systems, especially catastrophic weather systems. This study applies the blown-up theory to analyze and predict a rainstorm case in Jiangsu Province of East China to explore the applicability of the blown-up theory. At the same time, a numerical simulation experiment is conducted using the Weather Research and Forecasting Model (WRF) v4.2. The numerical results are compared with the European Center for Medium Weather Forecasting (ECMWF) Reanalysis v5 (ERA5) data and the China Meteorological Administration (CMA) Land Data Assimilation System (CLDAS) products. It is found that there is a deviation in the simulation for the precipitation center, and further analysis indicates that it is likely related to the position of the simulated low-level shear line. On the other hand, the blown-up analyses are consistent with the actual situation and provide additional information besides the numerical simulation results. These results indicate that the blown-up charts and V-3θ diagrams are able to predict the weather system transformation, the rainfall area, and the evolution of the rainstorm, which confirms the applicability of the blown-up theory to rainstorm forecasts. This provides an auxiliary analysis method in addition to numerical simulations for rainstorm forecasts.

“Regions” as brands: an empirical validation of consumer-based brand equity framework and demonstration of cross-over effect

PurposeThis study aims to empirically validate the holistic consumer-based brand equity (CBBE) framework and establishment of the cross-over effect using consumer preferences for wine “regions.”Design/methodology/approachThe authors utilize a sample of 275 regular wine purchasers in Australia, who were aged above 30 years old and had consumed regional wine in the last 3 months through an online consumer panel survey.FindingsBy testing competing CBBE models, the main research model was found to have the greatest predictive ability, due to its inclusion of both the rational and emotional paths and the cross-over effect between consumer brand judgment and brand feeling. This paper indicates that consumers enhance brand equity and develop a strong ongoing relationship with regional wine brands by simultaneously engaging in both mental behavior paths, where, in the meantime, rational processing has a cross-over effect on emotional processing.Originality/valueThis study expands the brand management literature by answering the future research on the CBBE model identified by Hall et al. (2021). Moreover, it establishes the underexplored cross-over effect within the CBBE literature. Furthermore, it adds to the marketing and wine-marketing literature by extending the concept of the brand to other attributes in consumer choice, such as a wine region as outlined by Giacomarra et al. (2020). Thus, this study advances the existing branding knowledge in a practical sense, which enables regional wine marketers and wine retailers to undertake promotional and product development strategies accordingly.

Injury severity prediction of cyclist crashes using random forests and random parameters logit models

Cycling provides numerous benefits to individuals and to society but the burden of road traffic injuries and fatalities is disproportionately sustained by cyclists. Without awareness of the contributory factors of cyclist death and injury, the capability to implement context-specific and appropriate measures is severely limited. In this paper, we investigated the effects of the characteristics related to the road, the environment, the vehicle involved, the driver, and the cyclist on severity of crashes involving cyclists analysing 72,363 crashes that occurred in Great Britain in the period 2016–2018. Both a machine learning method, as the Random Forest (RF), and an econometric model, as the Random Parameters Logit Model (RPLM), were implemented.Three different RF algorithms were performed, namely the traditional RF, the Weighted Subspace RF, and the Random Survival Forest. The latter demonstrated superior predictive performances both in terms of F-measure and G-mean. The main result of the Random Survival Forest is the variable importance that provides a ranked list of the predictors associated with the fatal and severe cyclist crashes. For fatal classification, 19 variables showed a normalized importance higher than 5% with the second involved vehicle manoeuvring and the gender of the driver of the second vehicle having the greatest predictive ability. For serious injury classification, 13 variables showed a normalized importance higher than 5% with the bike leaving the carriageway having the greatest normalized importance. Furthermore, each path from the root node to the leaf nodes has been retraced the way back generating 361 if-then rules with fatal crash as consequent and 349 if-then rules with serious injury crash as consequent. The RPLM showed significant unobserved heterogeneity in the data finding four normal distributed indicator variables with random parameters: cyclist age ≥ 75 (fatal prediction), cyclist gender male (fatal and serious prediction), and driver aged 55–64 (serious prediction). The model’s McFadden Pseudo R2 is equal to 0.21, indicating a very good fit. Furthermore, to understand the magnitude of the effects and the contribution of each variable to injury severity probabilities the pseudo-elasticity was assessed, gaining valuable insights into the relative importance and influence of the variables.The RF and the RPLM resulted complementary in identifying several roadways, environmental, vehicle, driver, and cyclist-related factors associated with higher crash severity. Based on the identified contributory factors, safety countermeasures useful to develop strategies for making bike a safer and more friendly form of transport were recommended.

Prognostic value of nutritional and inflammatory markers in patients with hepatocellular carcinoma who receive immune checkpoint inhibitors.

The emergence of immune checkpoint inhibitors (ICIs) has provided a new treatment option for patients with hepatocellular carcinoma (HCC). However, further evaluation is needed for determining biomarkers for the use of ICIs. The present study evaluated the prognostic value of certain nutritional and inflammatory markers in patients with HCC who received ICIs. In the present study, the clinical data of 151 patients with HCC who received ICIs at Harbin Medical University Cancer Hospital from January 2019 to December 2021 were collected. The blood parameters of all patients before treatment were collected to evaluate certain nutritional and inflammatory markers, including the prognostic nutrition index (PNI), nutritional risk index (NRI), geriatric NRI (GNRI), systemic immune-inflammation index (SII), systemic inflammation response index (SIRI) and advanced lung cancer inflammation index (ALI). Patients were grouped using the cut-off value calculated using receiver operating characteristic (ROC) curves, and the relationship between these biomarkers and prognosis was evaluated through survival analysis. Furthermore, the prognostic value of these biomarkers was assessed through multivariate Cox regression analysis and construction of nomograms. Finally, time-ROC curves were plotted to compare the differences in predicting prognosis between the biomarkers. In the preliminary survival analysis, all inflammatory and nutritional markers included in the present study were significantly associated with the prognosis of HCC in patients who received ICIs. Similar results were obtained in a subgroup analysis of patients with different Barcelona Clinic Liver Cancer (BCLC) stages. Multivariate Cox regression analysis demonstrated that GNRI, PNI, BCLC stage and Tumor-Node-Metastasis (TNM) stage were significantly associated with progression-free survival (PFS), whereas GNRI, BCLC stage and TNM stage were also significantly associated with overall survival (OS). Furthermore, the time-ROC curves indicated that nutritional indicators had a higher prognostic value in all indexes, especially GNRI. The C-index (95% confidence interval) of the nomograms for predicting the survival probability of patients who received ICIs were 0.801 (0.746-0.877) and 0.823 (0.761-0.898) for PFS and overall OS, respectively, which also showed high accuracy. In conclusion, the present study demonstrated that PNI, GNRI, NRI, SII, SIRI and ALI were all related to the efficacy of ICIs in HCC and could serve as non-invasive biomarkers for ICI treatment effectiveness. Moreover, compared with inflammatory markers, nutritional markers had greater predictive ability, with GNRI being the biomarker with the best prognostic value.

A data-based reduced-order model for dynamic simulation and control of district-heating networks

This study concerns the development of a data-based compact model for the prediction of the fluid temperature evolution in district heating (DH) pipeline networks. This so-called “reduced-order model” (ROM) is obtained from reduction of the conservation law for energy for each pipe segment to a semi-analytical input–output relation between the pipe outlet temperature and the pipe inlet and ground temperatures that can be identified from training data. The ROM basically is valid for generic pipe configurations involving 3D unsteady heat transfer and 3D steady flow as long as heat-transfer mechanisms are linearly dependent on the temperature field. Moreover, the training data can be generated by physics-based computational “full-order” models (FOMs) yet also by (calibration) experiments or field measurements. Performance tests using computational training data for a single-pipe configuration demonstrate that the ROM (i) can be successfully identified and (ii) can accurately describe the response of the outlet temperature to arbitrary input profiles for inlet and ground temperatures. Application of the ROM to two case studies, i.e. fast simulation of a small DH network and design of a controller for user-defined temperature regulation of a DH system, demonstrate its predictive ability and efficiency also for realistic systems. Dedicated cost analyses further reveal that the ROM may significantly reduce the computational costs compared to FOMs by (up to) orders of magnitude for higher-dimensional pipe configurations. These findings advance the proposed ROM as a robust and efficient simulation tool for practical DH systems with a far greater predictive ability than existing compact models.

Classification of Signature-Based Phenotypes of Aging-Related Genes to Identify Prognostic and Immune Characteristics in HCC.

Hepatocellular carcinoma (HCC), which has become one of the most significant malignancies causing cancer-related mortality, presents genetic and phenotypic heterogeneity that makes predicting prognosis challenging. Aging-related genes have been increasingly reported as significant risk factors for many kinds of malignancies, including HCC. In this study, we comprehensively dissected the features of transcriptional aging-relevant genes in HCC from multiple perspectives. We applied public databases and self-consistent clustering analysis to classify patients into C1, C2, and C3 clusters. The C1 cluster had the shortest overall survival time and advanced pathological features. Least absolute shrinkage and selection operator (LASSO) regression analysis was adopted to build the prognostic prediction model based on six aging-related genes (HMMR, S100A9, SPP1, CYP2C9, CFHR3, and RAMP3). These genes were differently expressed in HepG2 cell lines compared with LO2 cell lines measured by the mRNA expression level. The high-risk score group had significantly more immune checkpoint genes, higher tumor immune dysfunction and exclusion score, and stronger chemotherapy response. The results indicated that the age-related genes have a close correlation with HCC prognosis and immune characteristics. Overall, the model based on six aging-associated genes demonstrated great prognostic prediction ability.

A novel nomogram for identifying candidates for adjuvant chemotherapy in patients with stage IB gastric adenocarcinoma

BackgroundThe purpose of this research was to construct a novel predictive nomogram to identify specific stage IB gastric adenocarcinoma (GAC) populations who could benefit from postoperative adjuvant chemotherapy (ACT).MethodBetween 2004 and 2015, 1889 stage IB GAC patients were extracted from the Surveillance, Epidemiology, and End Results (SEER) program database. Then Kaplan–Meier survival analysis, univariate and multivariable Cox analyses, and univariate and multivariable logistic analyses were implemented. Finally, the predictive nomograms were constructed. The methods of area under the curve (AUC), calibration curve, and decision curve analysis (DCA) were used to validate the clinical effectiveness of the models.ResultsOf these patients, 708 cases underwent ACT, while the other 1181 patients didn’t receive ACT. After PSM, the patients in the ACT group presented a longer median overall survival (133 vs. 85 months, p = 0.0087). Among the ACT group, 194 (36.0%) patients achieving more prolonged overall survival than 85 months were regarded as the beneficiary population. Then the logistic regression analyses were performed, and age, gender, marital status, primary site, tumor size, and regional nodes examined were included as predicting factors to construct the nomogram. The AUC value was 0.725 in the training cohort and 0.739 in the validation cohort, which demonstrated good discrimination. And calibration curves indicated ideal consistency between the predicted and observed probabilities. Decision curve analysis presented a clinically useful model. Furthermore, the prognostic nomogram predicting 1-, 3-, and 5-year cancer-specific survival presented good predictive ability.ConclusionThe benefit nomogram could guide clinicians in decision-making and selecting optimal candidates for ACT among stage IB GAC patients. And the prognostic nomogram presented great prediction ability for these patients.

Validation of the Domestic High Bleeding Risk Criteria for Japanese Patients with Acute Myocardial Infarction.

The Academic Research Consortium (ARC) has proposed international criteria to standardize the definition of high bleeding risk (HBR) in patients undergoing percutaneous coronary intervention (PCI). In this context, Japan has also established its own guidelines, that is, the Japanese version of HBR (J-HBR) criteria. However, the J-HBR criteria have not been fully validated, especially in patients with acute myocardial infarction (MI). This bi-center registry included 1079 patients with acute MI undergoing primary PCI in a contemporary setting. Patient bleeding risks were evaluated using the ARC-HBR and J-HBR criteria. The primary endpoint was rates of major bleeding events (Bleeding Academic Research Consortium type 3 or 5) at 1 year. Of the 1079 patients, 505 (46.8%) and 563 (52.2%) met the ARC-HBR and J-HBR criteria, respectively. Patients who met the J-HBR criteria were found to have a higher rate of major bleeding events at 1 year than those who did not (12.8% vs. 3.3%, p＜0.001). When patients were scored and stratified using the J-HBR major and minor criteria, risks of major bleedings were progressively increased with the increase in the number of J-HBR criteria. In the receiver operating characteristic curve analysis, the ARC-HBR and J-HBR significantly predicted subsequent major bleedings after PCI, with ARC-HBR having greater predictive ability than J-HBR. More than half of the patients with acute MI undergoing primary PCI in Japan met the J-HBR criteria. Although the J-HBR criteria successfully identified patients who were likely to develop major bleeding events after primary PCI, the superiority of J-HBR to ARC-HBR in predicting bleeding outcomes warrants further investigation.

Association between aberrant amino acid metabolism and nonchromosomal modifications fetal structural anomalies: A cohort study.

More than half of the cases of fetal structural anomalies have no known cause with standard investigations like karyotype testing and chromosomal microarray. The differential metabolic profiles of amniotic fluid (AF) and maternal blood may reveal valuable information about the physiological processes of fetal development, which may provide valuable biomarkers for fetal health diagnostics. This cohort study of singleton-pregnant women had indications for amniocentesis, including structural anomalies and a positive result from maternal serum screening or non-invasive prenatal testing, but did not have any positive abnormal karyotype or chromosomal microarray analysis results. A total of 1580 participants were enrolled between June 2021 and March 2022. Of the 1580 pregnant women who underwent amniocentesis, 294 were included in the analysis. There were 137 pregnant women in the discovery cohort and 157 in the validation cohort. High-coverage untargeted metabolomic analysis of AF revealed distinct metabolic signatures with 321 of the 602 metabolites measured (53%) (false discovery rate, q < 0.005), among which amino acids predominantly changed in structural anomalies. Targeted metabolomics identified glutamate and glutamine as novel predictive markers for structural anomalies, their vital role was also confirmed in the validation cohort with great predictive ability, and the area under the receiver operating characteristic curves (AUCs) were 0.862 and 0.894 respectively. And AUCs for glutamine/glutamate were 0.913 and 0.903 among the two cohorts. Our results suggested that the aberrant glutamine/glutamate metabolism in AF is associated with nonchromosomal modificantions fetal structural anomalies. Based on our findings, a novel screening method could be established for the nonchromosomal modificantions fetal structural anomalies. And the results also indicate that monitoring fetal metabolic conditions (especially glutamine and glutamine metabolism) may be helpful for antenatal diagnosis and therapy.

Toward improved prediction of recycled brick aggregate concrete compressive strength by designing ensemble machine learning models

The utilization of recycled brick aggregate concrete (RBC) is an area of active research, in which further investigation is needed to develop accurate models for predicting the behavior of RBC to ensure its safe and sustainable use in construction. This study presents an effective way to determine RBC's compressive strength (CS) based on an appropriate design of ensemble machine learning (ML) models, namely Gradient Boosting, Light Gradient boosting, AdaBoost, Extreme Gradient Boosting, Stacking, and Voting. A database covering 393 test results is compiled from the relevant literature for training and testing the models. In addition, 10-fold cross-validation and random splitting of data are used to ensure the reliability of predictions, as well as prevent overfitting. The findings reveal that the Stacking model has the greatest predictive ability, with a coefficient of determination of 0.95, root mean square error of 2.74 MPa, mean absolute error of 2.09 MPa, mean absolute percentage error of 0.10 on the testing dataset. In addition, Feature importance and Partial dependence plots analysis are utilized to investigate the impact of each RBC component on its CS. In RBC’s mixed-component design, the outcomes of this research, along with a developed Graphical User Interface (GUI) for the CS, might be of great use to material engineers.

A promising subgroup identification method based on a genetic algorithm for censored survival data

ABSTRACT Modern precision medicine requires drug development to account for patients’ heterogeneity, as only a subgroup of the patient population is likely to benefit from the targeted therapy. In this paper, we propose a novel method for subgroup identification based on a genetic algorithm. The proposed method can detect promising subgroups defined by predictive biomarkers in which the treatment effects are much higher than the population average. The main idea is to search for the subgroup with the greatest predictive ability in the entire subgroup space via a genetic algorithm. We design a real-valued representation of subgroups that evolves according to a genetic algorithm and derive an objective function that properly evaluates the predictive ability of the subgroups. Compared with model- or tree-based subgroup identification methods, the distinctive search strategy of this new approach offers an improved capability to explore subgroups defined by multiple predictive biomarkers. By embedding a resampling scheme, the multiplicity and complexity issues inherent in subgroup identification methods can be addressed flexibly. We evaluate the performance of the proposed method in comparison with two other methods using simulation studies and a real-world example. The results show that the proposed method exhibits good properties in terms of multiplicity and complexity control, and the subgroups identified are much more accurate. Although we focus on the implementation of censored survival data, this method could easily be extended for the realization of continuous and categorical endpoints.