Development Of Predictive Models Research Articles

Investigation and prediction of fatigue performance of SLM 316 L stainless steel based on small build orientation variations and heat treatment effects

Additive manufacturing (3D printing) has facilitated the creation of complex structures, including lattices with diverse build angles. Since lattices often face quasi-static and cyclic loads, understanding material properties across a wide range of printing angles, post-processing conditions, and loading scenarios is crucial. This study explores the effects of building orientations beyond conventional 0, 45, and 90 degrees on the mechanical properties of 316 L steel, produced by laser powder bed fusion. Quasi-static and cyclic tensile tests were performed to evaluate mechanical behavior in as-built and heat-treated conditions, followed by predictive model development. The ultimate strength showed a modest variation of 7% in the as-built condition, while the heat-treated state exhibited a greater variation of 13%. Fatigue tests indicated minor differences between conditions in the low-cycle region but larger gaps in the high-cycle region, where heat-treated samples generally showed superior performance. Inclination angle had a greater effect on fatigue life in the as-built state, with horizontal part orientations outperforming vertical ones. The predictive model demonstrated robust reliability, with nearly 90% of data within an accepted scatter factor of less than three. Requiring minimal experimental data (two fatigue curves per condition), the model is valuable for forecasting fatigue behavior in complex lattice structures.

Cross-contextual stress prediction: Simple methodology for comparing features and sample domain adaptation techniques in vital sign analysis

Stress significantly impacts individuals, particularly in professions like nursing and driving, leading to severe health risks and accidents. Accurate stress measurement is critical for effective interventions, yet research is hindered by incomplete datasets and inconsistent methodologies, slowing the development of reliable predictive models. This paper introduces a framework for cross-contextual stress prediction, enabling the generation of general stress prediction models adaptable to specific domain challenges. The methodology leverages two general daily life datasets and three domain-specific datasets, employing steps such as dataset selection, feature extraction, significant feature identification, feature preprocessing, fine-tuning, domain adaptation, and application to specific contexts. Through this framework, key vital signs were identified as significant predictors of stress, including electrocardiography (ECG), heart rate (HR), heart rate variability (HRV) - low frequency (LF), electrodermal activity (EDA), body temperature (TEMP), and skin conductance response (SCR). The experiments conducted include: 1) Utilizing HR and HRV-LF through domain adaptation from general to automobile driving datasets; 2) Applying EDA, HR, and TEMP from general to specific nurse activity datasets; and 3) Adapting ECG, HR, and TEMP from general to automobile driving datasets. Results demonstrate the potential of the proposed framework for cross-contextual stress prediction, with HR and HRV-LF identified as pivotal features. When applied to target datasets specific to stress scenarios, the model achieved a 62% F1 score, demonstrating the effectiveness of the feature-based Correlation Alignment (CORAL) technique combined with Random Forest models in transferring learned knowledge across domains. These findings highlight the robustness of the approach in adapting general stress prediction models to specific contexts, paving the way for real-world applications such as stress monitoring in driving and nursing during high-stress periods like COVID-19.

Development and validation of a risk prediction model for PICC-related venous thrombosis in patients with cancer: a prospective cohort study

To develop and validate a risk prediction model for predicting the risk of Peripherally Inserted Central Catheter-Related venous thrombosis (PICC-RVT) in cancer patients with PICCs. A prospective cohort study of 281 cancer patients with PICCs was conducted from April 2023 to January 2024. Data on patient-, laboratory- and catheter-related risk factors were collected on the day of catheterization. Patients were investigated for PICC-RVT by Doppler sonography in the presence of PICC-RVT signs and symptoms. Univariate and multivariate regression analyses were used to identify independently associated risk factors of PICC-RVT and develop a risk prediction model. 275 patients were finally included for data analysis, and 18 (6.5%) developed PICC-RVT. Four risk factors were identified as key predictors of PICC-RVT, including “diabetes requiring insulin (OR: 8.016; 95% CI 1.157–55.536), major surgery (within 1 month and operation time > 45 minutes) (OR: 0.023; 95% CI 1.296–30.77), reduced limb activities of the PICC arm (OR: 6.687; 95% CI 2.024–22.09)” and “catheter material (OR: 3.319; 95% CI 0.940–11.723)”. The nomogram model was developed and internally validated with an area under the receiver operating characteristics curve (AUC) of 0.796 (95% CI 0.707–0.885). The Hosmer–Lemeshow goodness-of-ft was 1.685 (p = 0.194). The nomogram prediction model had good predictive performance. This model could help identify patients at the highest risk for PICC-RVT to guide effective prophylaxis. Further external validation studies of this nomogram model on a large sample are required.

Development and validation of a deep learning-enhanced prediction model for the likelihood of pulmonary embolism

BackgroundPulmonary embolism (PE) is a common and potentially fatal condition. Timely and accurate risk assessment in patients with acute deep vein thrombosis (DVT) is crucial. This study aims to develop a deep learning-based, precise, and efficient PE risk prediction model (PE-Mind) to overcome the limitations of current clinical tools and provide a more targeted risk evaluation solution.MethodsWe analyzed clinical data from patients by first simplifying and organizing the collected features. From these, 37 key clinical features were selected based on their importance. These features were categorized and analyzed to identify potential relationships. Our prediction model uses a convolutional neural network (CNN), enhanced with three custom-designed modules for better performance. To validate its effectiveness, we compared this model with five commonly used prediction models.ResultsPE-Mind demonstrated the highest accuracy and reliability, achieving 0.7826 accuracy and an area under the receiver operating characteristic curve of 0.8641 on the prospective test set, surpassing other models. Based on this, we have also developed a Web server, PulmoRiskAI, for real-time clinician operation.ConclusionThe PE-Mind model improves prediction accuracy and reliability for assessing PE risk in acute DVT patients. Its convolutional architecture and residual modules substantially enhance predictive performance.

Integrating Public Health Data and Epidemiological Models in Business Decision-Making Post COVID-19

Objective: Propose the integration of public health data and epidemiological models into decision-making processes for health-related businesses in the Panamanian economy during future pandemics. Theoretical Framework: The analysis of the relationship between public health and the economy, through econometric models, basic reproduction number, effective reproduction number, and their interaction in health crisis contexts, provides a foundation for understanding the use of epidemiological data in strategic crisis management. Method: The methodology involves a quantitative approach using time series and predictive models applied to secondary data from the Ministry of Health of Panama on COVID-19 cases between march-april 2020. Results and Discussion: A strong relationship was found between hospitalized cases in regular wards and ICUs, consistent with findings from Kucharski et al. (2020) and Petrucci et al. (2020). This suggests that an increase in direct infections translates to higher demand for hospital services, enabling the anticipation of risks and strategic planning for staff and supplies. Research Implications: The implementation of models that account for trends, seasonality, cycles, and irregularities, complemented by the analysis of relationships between variables, facilitates the development of predictive models that incorporate epidemiological indicators and their effects on any economic sector, particularly health-related ones. Originality/Value: This study proposes a comprehensive framework that combines public health data with econometric models, offering an innovative tool for business decision-making during health crises.

Individualized drug therapy and survival prediction in ICU patients with acute kidney injury: construction and validation of a nomogram

BackgroundAcute kidney injury (AKI) is defined by a sharp decrease in the estimated glomerular filtration rate (eGFR). However, the impact of medication history on the survival of AKI patients has received little attention. Hence, it is necessary to investigate the potential of medication history as a predictor of survival outcomes among AKI patients in the intensive care unit (ICU).MethodsCritically ill AKI patients were sourced from the MIMIC-IV database. To ascertain significant, drug-related, independent predictors of survival, univariate Cox analysis and stepwise Cox regression were performed. Based on the identified predictor, a nomogram was developed to estimate the individualized survival probability for AKI patients. Additionally, to address potential confounders among patients with medications referenced in the nomogram, a propensity score matching procedure was applied. Ultimately, a comparative analysis was performed to elucidate the prognostic disparities among these patient subgroups.ResultsThis study enrolled 1,208 patients and developed a nomogram incorporating oxygen flow rate, respiratory frequency, continuous venovenous hemodiafiltration status, age, and medication use (including ibuprofen, epinephrine, cefazolin, warfarin, and vasopressin). The predictive model demonstrated diagnostic accuracy, with AUC values for 1-year, 3-year, and 5-year survival among AKI patients of 0.827, 0.799, and 0.777 in the training dataset, and 0.760, 0.743, and 0.740 in the internal validation dataset, respectively. Kaplan–Meier survival analyses revealed significant differences in survival outcomes among AKI patients based on their exposure to different medications.ConclusionsIn summary, the developed prediction model demonstrated accuracy for AKI patients in the ICU and helped clinical decision-making. However, future studies will require external validation to confirm these findings.

Harnessing Deep Learning and Technical Indicators for Enhanced Stock Predictions of Blue-Chip Stocks on the Indonesia Stock Exchange (IDX)

Given the limitations of existing models in accurately predicting stock prices, particularly in emerging markets such as Indonesia, this study aimed to evaluate the effectiveness of deep learning models in forecasting stock prices using blue-chip company shares traded on the Indonesia Stock Exchange (IDX). The main focus lies in combining historical stock data with a series of existing technical indicators, optimizing their integration to improve prediction accuracy. The accuracy of this method is reflected in a comprehensive evaluation of model performance using robust metrics, including R2, Mean Squared Error (MSE), and Root Mean Squared Error (RMSE). Empirical results show the superiority of models integrating technical indicators compared to models relying only on historical data. The LSTM model showed the most significant improvement, with R2 for ASII stock jumping by 14.59% after incorporating technical indicators. The prediction accuracy of the GRU model for BBCA shares increased significantly, as shown by a decrease of 45.16% in MSE. These findings underscore the critical role of feature selection in developing prediction models. Integrating technical indicators with historical stock data increases prediction accuracy and provides additional tools for informed decision-making.

Experimental Study on the Flexural Performance of Geogrid-Reinforced Foamed Lightweight Soil

The flexural behavior of geogrid-reinforced foamed lightweight soil (GRFL soil) is investigated in this study using unconfined compressive and four-point bending tests. The effects of wet density and reinforcement layers on flexural performance are analyzed using load–displacement curves, damage patterns, load characteristics, unconfined compressive strength, and flexural strength. A variance study demonstrates that increasing the wet density significantly increases unconfined compressive strength. Bond stress mechanisms enable geogrid integration, efficiently reroute stresses internally, and greatly increase flexural strength. With a maximum unconfined compressive strength of 3.16 MPa and a peak flexural strength increase of 166%, this reinforcement increases both strength and ductility by changing the damage pattern from brittle to ductile. The principal load is initially supported by the foamed lightweight soil, and in later phases, geogrids take over load-bearing responsibilities. Additionally, the work correlates the ratio of unconfined compressive to flexural strength with wet density and informs the development of predictive models for unconfined compressive strength as a function of reinforcing layers and wet density.

Evaluating immune cell abundance across hepatocellular carcinoma disease stages I-III: Findings from TCGA analysis.

621 Background: Hepatocellular carcinoma (HCC), the predominant form of primary liver cancer, is the sixth most prevalent cancer globally and a leading cause of cancer-related mortality. The tumor microenvironment (TME) in HCC, characterized by complex interactions between immune and tumor cells, significantly influences tumor progression and immune suppression. Understanding the immune landscape across the different stages of HCC can inform prognosis and therapeutic strategies. Methods: We analyzed RNA seq data from 333 HCC patients (disease stages I-III), from The Cancer Genome Atlas (TCGA), using the immune deconvolution algorithms CIBERSORT and xCell. Differential gene expression analysis was performed using the DESeq2 package in R. Immune cell abundance across various HCC stages was imputed, and survival analyses was conducted using Kaplan-Meier curves and log-rank tests. Results: Of the 19,514 genes analyzed, 2,774 were up-regulated and 947 were down-regulated in stage II and III in comparison to stage I. CIBERSORT and xCell analyses revealed significant variability in the immune cell composition between HCC stages. Notable differences were observed in the T cells (CD4 and CD8 subsets), B cells, plasma cells, and stromal components. ANOVA and post hoc tests identified significant stage-specific differences in several immune cell types. For the survival analysis, we categorized patients into low and high groups based on median cell abundances. Overall survival showed no statistically significant differences among the various groups for common myeloid and lymphoid progenitor cells. In contrast, disease-free survival demonstrated statistically significant differences across the groups for both cell types. Conclusions: This study emphasizes the tumor microenvironment's pivotal role in HCC, revealing that immune cell composition significantly influences disease-free survival, despite limited effects on overall survival. These findings advocate for further exploration into targeted immunotherapies and the development of predictive models that integrate immune landscape characteristics, ultimately aiming to enhance patient outcomes in HCC. Cell type xCell Cibersort Mean SD Mean SD T cell CD4 naive 0.00466 0.01994 1.31e-5 1.69e-4 T cell CD8 0.01368 0.03253 0.03161 0.04947 Eosinophil 3.96e-5 3.41e-4 1.67e-4 8.69e-4 Macrophage 0.02323 0.02441 0.02924 0.04345 Macrophage.M1 0.00914 0.01526 0.01806 0.02054 Macrophage.M2 0.02303 0.01697 0.05439 0.04815 B cell memory 0.00196 0.01418 0.00107 0.00733 B cell naive 0.00163 0.01034 0.01319 0.02148 Neutrophil 2.14e-4 0.00137 0.00112 0.00539 Immune score 0.04621 0.06817 - - Table presents various cell type abundances for each algorithm used.

Initializing a public repository for hosting benchmark datasets to facilitate machine learning model development in food safety.

While there is clear potential for artificial intelligence (AI) and machine learning (ML) models to help improve food safety, the development and deployment of these models in the food safety domain is by and large lacking. The absence of publicly available databases that host well-curated datasets that can be used to develop and validate AI /ML models represents one likely barrier. Thus, we took three previously published datasets, which we further cleaned and annotated, and made them publicly available in a repository called Cornell Food Safety ML Repository. The selected datasets include (i) presence or absence of Listeria spp. in soil samples collected across the U.S. with paired metadata for soil properties, geolocation, climate, and surrounding land use, (ii) presence or absence of Salmonella and Campylobacter in young chicken carcasses tested in processing facilities with associated meteorological and temporal metadata, and (iii) presence or absence of fecal contamination as well as E. coli concentration in New York watersheds with associated metadata for land use, water attributes, and meteorological factors. These datasets can serve as benchmark datasets for developing ML models. To demonstrate the utility of the repository, we developed customizable scripts as well as LazyPredict (a quick screening method) scripts for training different types of ML models using the shared datasets. While this repository provides an important starting point that will allow for the development and testing of ML models to predict foodborne pathogens contamination in different sources, inclusion of further datasets is clearly needed to advance this field. This paper thus includes a call to action for the deposit of well-curated datasets that can be used for further development of predictive models in food safety. This paper will also discuss the benefits of such public databases, including the assessment of data-sharing scenarios using existing privacy-preserving techniques.

Outcomes Prognostication for Septuagenarians and Octogenarians Diagnosed with Meningioma: Leveraging National Registry Data for the Development of a Prognostic Model for Survival Prediction in the Elderly Population

Outcomes Prognostication for Septuagenarians and Octogenarians Diagnosed with Meningioma: Leveraging National Registry Data for the Development of a Prognostic Model for Survival Prediction in the Elderly Population

Development and validation of an explainable machine learning model for mortality prediction among patients with infected pancreatic necrosis.

Infected pancreatic necrosis (IPN) represents a severe complication of acute pancreatitis, commonly linked with mortality rates ranging from 15% to 35%. However, the present mortality prediction tools for IPN are limited and lack sufficient sensitivity and specificity. This study aims to develop and validate an explainable machine learning (ML) model for death prediction among patients with IPN. We performed a prospective cohort study of 344 patients with IPN consecutively enrolled from a large Chinese tertiary hospital from January 2011 to January 2023. Ten ML models were developed to predict 90-day mortality in these patients. A benchmarking test, involving nested resampling, automatic hyperparameter tuning and random search techniques, was conducted to select the ML model. Sequential forward selection method was employed to select the optimal feature subset from 31 candidate subsets to simplify the model and maximize predictive performance. The final model was internally validated with the 1000 bootstrap method and externally validated using an independent cohort of 132 patients with IPN retrospectively collected from another Chinese tertiary hospital from January 2018 to January 2023. The SHapley Additive exPlanations (SHAP) method was employed to interpret the model in terms of features importance and features effect. The final model constructed with optimal feature subset was deployed as an interactive web-based Shiny app. Random survival forest (RSF) model showed the best predictive performance than other 9ML models (internal validation, C-index=0.863 [95% CI: 0.854-0.875]; external validation, C-index=0.857 [95% CI: 0.850-0.865]). Multiple organ failure, Acute Physiology and Chronic Health Examination II (APACHE II) score ≥20, duration of organ failure ≥21 days, bloodstream infection, time from onset to first intervention <30 days, Bedside Index of Severity in Acute Pancreatitis score ≥3, critical acute pancreatitis, age ≥ 50 years, and hemorrhage were 9 most important features associated with mortality. Furthermore, SHAP algorithm revealed insightful nonlinear interactive associations between important predictors and mortality, identifying 9 features pairs with high interaction SHAP value and clinical significance. Two interactive web-based Shiny apps were developed to enhance clinical practicability: https://rsfmodels.shinyapps.io/IPN_app/ for cases where the APACHE II score was available and https://rsfmodels.shinyapps.io/IPNeasy/ for cases where it was not. An explainable ML model for death prediction among IPN patients was feasible and effective, suggesting its superior potential in guiding clinical management and improving patient outcomes. Two publicly accessible web tools generated for the optimized model facilitated its utility in clinical settings. The Natural Science Foundation of Hunan Province (2023JJ30885), Postdoctoral Fellowship Program of CPSF (GZB20230872), The Youth Science Foundation of Xiangya Hospital (2023Q13), The Project Program of National Clinical Research Center for Geriatric Disorders of Xiangya Hospital (2021LNJJ19).

Development of a nomogram model for early prediction of refractory convulsive status epilepticus.

We aim to identify risk factors that predict refractory convulsive status epilepticus (RCSE) and to develop a model for early recognition of patients at high risk for RCSE. This study involved 200 patients diagnosed with convulsive status epilepticus (CSE), of whom 73 were RCSE and 127 were non-RCSE. Variables included demographic information, lifestyle factors, medical history, comorbidities, clinical symptoms, neuroimaging characteristics, laboratory tests, and nutritional scores. A predictive model was developed through multivariable logistic regression analysis. The model's predictive performance and clinical utility were evaluated using various metrics, including the area under the receiver operating characteristic (AUROC) curve, GiViTI calibration belt, and decision curve analysis (DCA). Additionally, we performed internal five-fold cross-validation for this model. We developed a nomogram model with six predictors: age≤40years, prior history of epilepsy, presence of epileptic foci, duration of CSE>30min, c-reactive protein>6mg/L, and nutritional risk screening≥3 points. Our model has a high AUROC (0.838) and good consistency (P = 0.999). In DCA, the curve of our model exhibits a positive net benefit across the entire range of threshold probabilities. Moreover, our model achieved an accuracy of 0.778 and a Kappa value of 0.519 in the five-fold cross-validation. We developed an objective, simple and accessible model to assess the risk of RCSE. This model shows promise as a valuable tool for evaluating the individual risk of RCSE.

Integrating deep learning algorithms for forecasting evapotranspiration and assessing crop water stress in agricultural water management.

The increasing impacts of climate change on global agriculture necessitate the development of advanced predictive models for efficient water management in crop fields. This study aims to enhance the forecasting of evapotranspiration (ET), potential evapotranspiration (PET), and crop water stress index (CWSI) using state-of-the-art deep learning techniques. This research integrates high-resolution climatic data from the ACCESS-ESM model and incorporates four shared socioeconomic pathways (SSPs) to represent a wide range of future climate scenarios. We employ feed forward neural networks (FFNNs), convolutional neural networks (CNNs), gated recurrent units (GRUs), and long short-term memory networks (LSTMs) to predict ET, PET, and CWSI. These findings reveal significant improvements in prediction accuracy, offering valuable insights for agricultural water management in Bangladesh. This approach provides a robust framework for optimizing irrigation practices and enhancing crop resilience against climate-induced water stress.

Development and Validation of Prediction Models for Perioperative Opioid Requirements: Integrating Machine-Learning Approach with Conventional Regression Method

Development and Validation of Prediction Models for Perioperative Opioid Requirements: Integrating Machine-Learning Approach with Conventional Regression Method

Development and validation of non-invasive prediction models for assessing kidney histopathological activity index in lupus nephritis.

To develop and validate prediction models for estimating the activity index (AI) of kidney histopathology in lupus nephritis (LN) using clinical and laboratory data. This study used single-center data from patients with kidney biopsy-confirmed LN between January 2012 and March 2022. The development and validation datasets were temporally cut. We discriminated AI > 10 and ≤ 10 as high and mild/moderate activity status, respectively. We constructed four models for AI: Model 1 included demographic information; Model 2 additionally incorporated data of systemic conditions; Model 3 further included kidney-specific conditions; and Model 4 included all the aforementioned predictors. Logistic regression was employed in Models 1 to 3, while Model 4 utilized least absolute shrinkage and selection operator for predictor selection and model building. Internal validation was performed using 1000 bootstrap resampling, while external validation was performed in the temporal validation dataset. Both calibration and discrimination metrics were evaluated. There were 160 patients in the development dataset and 70 patients in the validation dataset. In the temporal validation, all the models achieved acceptable calibration and excellent discrimination. Model 2 which contained relatively fewer predictors achieved the highest area under the receiver operator characteristic curve of 0.86 (95% confidence interval 0.76 to 0.94). Our Model 2 incorporating demographic and systemic indicators exhibited good performance in estimating the AI of LN. We thus provide a simple yet effective algorithm to predict AI in patients with LN, potentially aiding clinicians in non-invasively assessing disease activity and guiding treatment decisions. Key Points • We developed a prediction model (Model 2) incorporating demographic and systemic indicators to predict AI in patients with LN. • The prediction model can aid clinicians in noninvasively assessing disease activity and guiding treatment decisions.

Development and validation of clinical prediction model for functional independence measure following stroke rehabilitation.

To develop and internally validate a clinical prediction model that includes balance ability and nutritional indices for the motor-functional independence measure (M-FIM) at 90 days post-stroke stroke. This retrospective, single-center study included 566 patients with stroke undergoing rehabilitation at our rehabilitation hospital. The primary outcome was the M-FIM score of >61 at 3 months post-strokes onset. Stepwise conditional forward selection was first used to identify predictors for the achievement of M-FIM>61 at 90 days post-stroke, from 25 potential predictors at admission. The selected predictors were dichotomized with cut-off values to establish scoring systems, resulting in the B-ADL model, which includes postural balance (B), albumin level, age, arm function (A), days since stroke onset (D), and level of activities of daily living (ADL) (L). For internal validation, we corrected the optimism of the area under the curve of receiver operating characteristic curve (AUROC) induced by overfitting the original data using the bootstrap validation method. Calibration capacity was assessed using a calibration plot. We developed a clinical model to predict the M-FIM at 90 days post-stroke onset. The AUROC of the B-ADL model was 0.92 (sensitivity, 93.7%; specificity, 89.7%). The B-ADL model showed high accuracy with an AUROC of 0.970 in the internal validation. The scoring system in the validation cohort had a cut-off value of 5.5/12 points to predict the achievement of M-FIM>61 (AUROC: 0.950; 95% CI 0.930-0.970). The B-ADL model accurately predicted M-FIM >61 at 90 days post-stroke on the day of admission to the recovery rehabilitation ward. The B-ADL model is useful for optimizing rehabilitation programs and resource allocation, allowing for targeted interventions after stroke.

Development and validation of a prognostic prediction model for elderly gastric cancer patients based on oxidative stress biochemical markers

BackgroundThe potential of the application of artificial intelligence and biochemical markers of oxidative stress to predict the prognosis of older patients with gastric cancer (GC) remains unclear.MethodsThis retrospective multicenter study included consecutive patients with GC aged ≥ 65 years treated between January 2012 and April 2018. The patients were allocated into three cohorts (training, internal, and external validation). The GC-Integrated Oxidative Stress Score (GIOSS) was developed using Cox regression to correlate biochemical markers with patient prognosis. Predictive models for five-year overall survival (OS) were constructed using random forest (RF), decision tree (DT), and support vector machine (SVM) methods, and validated using area under the curve (AUC) and calibration plots. The SHapley Additive exPlanations (SHAP) method was used for model interpretation.ResultsThis study included a total of 1,859 older patients. The results demonstrated that a low GIOSS was a predictor of poor prognosis. RF was the most efficient method, with AUCs of 0.999, 0.869, and 0.796 in the training, internal validation, and external validation sets, respectively. The DT and SVM models showed low AUC values. Calibration and decision curve analyses demonstrated the considerable clinical usefulness of the RF model. The SHAP results identified pN, pT, perineural invasion, tumor size, and GIOSS as key predictive features. An online web calculator was constructed based on the best model.ConclusionsIncorporating the GIOSS, the RF model effectively predicts postoperative OS in older patients with GC and is a robust prognostic tool. Our findings emphasize the importance of oxidative stress in cancer prognosis and provide a pathway for improved management of GC.Trial registrationRetrospectively registered at ClinicalTrials.gov (trial registration number: NCT06208046, date of registration: 2024–05-01).

Development of a Clinical Prediction Model to Predict Malignant Transformation of Sinonasal Inverted Papilloma Based on Hematological Indices and Clinical Features.

Sinonasal inverted papilloma (SNIP) has a risk of malignant transformation into SNIP with squamous cell carcinoma (SNIP-SCC). Early detection of SNIP-SCC is crucial for patient outcomes. This study aimed to evaluate the utility of tumor-related hematological indices in the early diagnosis of SNIP-SCC and to develop a nomogram incorporating imaging and clinical features. This study included 159 patients with SNIP with (n = 34) or without (n = 125) SCC. Univariate and multivariate logistic regression analyses were employed to identify independent risk factors and develop nomogram. The diagnostic model was evaluated using receiver operating characteristics (ROC), standard, clinical decision, and clinical impact curves. Elevated serum cytokeratin fragment antigen 21-1 (CYFRA 21-1), the loss of convoluted cerebriform pattern, bone destruction, headache/facial pain, and epistaxis/blood-tinged mucus were identified as independent risk factors for SNIP-SCC. The optimal cut-off value for serum CYFRA 21-1 was 3.51 ng/mL, with the nomogram area under the ROC curve of 0.966 based on the above indicators. The findings suggest that CYFRA 21-1 is a promising diagnostic marker for SNIP-SCC. The nomogram incorporating serum CYFRA 21-1, imaging features, and clinical factors demonstrates strong clinical utility and can be a reference tool for clinical decision-making.

Harnessing temporal patterns in administrative patient data to predict risk of emergency hospital admission.

Unplanned hospital admissions are associated with worse patient outcomes and cause strain on health systems worldwide. Primary care electronic health records (EHRs) have successfully been used to create prediction models for emergency hospitalisation, but these approaches require a broad range of diagnostic, physiological, and laboratory values. In this study, we aimed to capture temporal patterns of patient activity from EHR data and evaluate their effectiveness in predicting emergency hospital admissions compared with conventional methods. In this retrospective observational study, we used the Secure Anonymised Information Linkage databank to extract temporal patterns of primary care activity from undifferentiated electronic health record timestamp data for 1·37 million patients in Wales aged 18-80 years with at least one recorded Read code between the years 2016 and 2018. Using Gaussian mixture modelling we grouped patients into distinct temporal clusters, performed a three-stage validation of our approach and calculated the risk of emergency hospital admission for each temporal cluster group. Finally, these temporal clusters were combined with five administrative variables and incorporated into four emergency hospital admission prediction models (logistic regression, naive Bayes, XGBoost, and multilayer perceptron [MLP]) and compared with a more traditional, but data-intensive, modelling technique. The primary outcome was emergency hospital admission as the next health-care event. Six distinct temporal cluster patterns of primary care EHR activity were identified, associated with varying risks of future emergency hospital admission risk. These patterns were visually interpretable, repeatable at a population-level, and clinically plausible. The best emergency hospital admission prediction model (MLP) achieved an area under the receiver operating characteristic (AUROC) of 0·82 and precision of 0·94 in regional cohorts. In external validation in regional cohorts, similar model performance was observed (AUROC 0·82 and precision 0·92). This model also matched the performance of a more complex model (extended feature model) requiring 33 clinical parameters (AUROC 0·82 vs 0·83; precision 0·94 vs 0·90) for the same task on the same dataset. We developed a novel machine learning pipeline that extracts interpretable temporal patterns from simple representations of EHR data and can be incorporated into emergency hospital admission predictors. This framework might enable more rapid development of parsimonious clinical prediction models. UKRI CDT in AI for Healthcare, UKRI Turing AI Fellowship, NIHR Imperial Biomedical Research Centre, and Research Capability Funding.