Purpose: The study aimed to create a machine learning model (MLM) to predict the stone-free status (SFS) of patients undergoing percutaneous nephrolithotomy (PCNL) and compare its performance to the S.T.O.N.E. and Guy’s stone scores. Patients and Methods: This is a retrospective study that included 320 PCNL patients. Pre-operative and post-operative variables were extracted and entered into three MLMs: RFC, SVM, and XGBoost. The methods used to assess the performance of each were mean bootstrap estimate, 10-fold cross-validation, classification report, and AUC. Each model was externally validated and evaluated by mean bootstrap estimate with CI, classification report, and AUC. Results: Out of the 320 patients who underwent PCNL, the SFS was found to be 69.4%. The RFC mean bootstrap estimate was 0.75 and 95% CI: [0.65– 0.85], 10-fold cross-validation of 0.744, an accuracy of 0.74, and AUC of 0.761. The XGBoost results were 0.74 [0.63– 0.85], 0.759, 0.72, and 0.769, respectively. The SVM results were 0.70 [0.60– 0.79], 0.725, 0.74, and 0.751, respectively. The AUC of Guy’s stone score and the S.T.O.N.E. score were 0.666 and 0.71, respectively. The RFC external validation set had a mean bootstrap estimate of 0.87 and 95% CI: [0.81– 0.92], an accuracy of 0.70, and an AUC of 0.795, While the XGBoost results were 0.84 [0.78– 0.91], 0.74, and 0.84, respectively. The SVM results were 0.86 [0.80– 0.91], 0.79, and 0.858, respectively. Conclusion: MLMs can be used with high accuracy in predicting SFS for patients undergoing PCNL. MLMs we utilized predicted the SFS with AUCs superior to those of GSS and S.T.O.N.E scores.

Glioblastoma (GBM) represents a profoundly aggressive and heterogeneous brain neoplasm linked to a bleak prognosis. Hypoxia, a common feature in GBM, has been linked to tumor progression and therapy resistance. In this study, we aimed to identify hypoxia-related differentially expressed genes (DEGs) and construct a prognostic signature for GBM patients using multi-omics analysis. Patient cohorts were collected from publicly available databases, including the Gene Expression Omnibus (GEO), the Chinese Glioma Genome Atlas (CGGA), and The Cancer Genome Atlas—Glioblastoma Multiforme (TCGA-GBM), to facilitate a comprehensive analysis. Hypoxia-related genes (HRGs) were obtained from the Molecular Signatures Database (MSigDB). Differential expression analysis revealed 41 hypoxia-related DEGs in GBM patients. A consensus clustering approach, utilizing these DEGs’ expression patterns, identified four distinct clusters, with cluster 1 showing significantly better overall survival. Machine learning techniques, including univariate Cox regression and LASSO regression, delineated a prognostic signature comprising six genes (ANXA1, CALD1, CP, IGFBP2, IGFBP5, and LOX). Multivariate Cox regression analysis substantiated the prognostic significance of a set of three optimal signature genes (CP, IGFBP2, and LOX). Using the hypoxia-related prognostic signature, patients were classified into high- and low-risk categories. Survival analysis demonstrated that the high-risk group exhibited inferior overall survival rates in comparison to the low-risk group. The prognostic signature showed good predictive performance, as indicated by the area under the curve (AUC) values for one-, three-, and five-year overall survival. Furthermore, functional enrichment analysis of the DEGs identified biological processes and pathways associated with hypoxia, providing insights into the underlying mechanisms of GBM. Delving into the tumor immune microenvironment, our analysis revealed correlations relating the hypoxia-related prognostic signature to the infiltration of immune cells in GBM. Overall, our study highlights the potential of a hypoxia-related prognostic signature as a valuable resource for forecasting the survival outcome of GBM patients. The multi-omics approach integrating bulk sequencing, single-cell analysis, and immune microenvironment assessment enhances our understanding of the intricate biology characterizing GBM, thereby potentially informing the tailored design of therapeutic interventions.

The study aimed to create a machine learning model (MLM) to predict the stone-free status (SFS) of patients undergoing percutaneous nephrolithotomy (PCNL), and compare its performance to the S.T.O.N.E. and Guy’s stone scores. This is a retrospective study that included 320 PCNL patients. The pre-operative and post-operative variables were extracted and entered into three MLMs: RFC, SVM, and XGBoost. The methods used to assess the performance of each MLM, were mean bootstrap estimate with CI, 10-fold cross-validation, classification report, and AUC. Each model was externally validated and evaluated by mean bootstrap estimate with CI, classification report, and AUC. Out of the 320 patients who underwent PCNL the SFS was found to be 69.4%. The RFC mean bootstrap estimate was 0.75 and 95% CI: [0.65–0.85], 10-fold cross-validation of 0.744, an accuracy of 0.74, and AUC of 0.761. The XGBoost results were 0.74 [0.63–0.85], 0.759, 0.72, and 0.769 respectively. The SVM results were 0.70 [0.60–0.79], 0.725, 0.74, and 0.751 respectively. The AUC of Guy’s stone score and the S.T.O.N.E. score were 0.666 and 0.71, respectively. The RFC external validation set had a mean bootstrap estimate of 0.87 and 95% CI: [0.81–0.92], an accuracy of 0.70, and an AUC of 0.795. While the XGBoost results were 0.84 [0.78–0.91], 0.74, and 0.84 respectively. The SVM results were 0.86 [0.80–0.91], 0.79, and 0.858 respectively. MLMs can be used with high accuracy in predicting SFS for patients undergoing PCNL. MLM systems we utilized predicted the SFS with AUCs superior to those of GSS and S.T.O.N.E score.

This study aimed to introduce a novel machine learning (ML) model designed to predict the need of interventions during endoscopy in patients with upper gastrointestinal bleeding (UGIB). The risk stratification tools in current use, such as the Glasgow Blatchford Score (GBS) and pre-endoscopic Rockall score, have limitations in accurately predicting the need for endoscopic interventions. All patients diagnosed with UGIB from January 2013 to October 2023 who underwent endoscopy were included in the study. Variables extracted included demographics, social history ,clinical history, clinical presentation and symptoms, drug history, management prior to endoscopy, findings during endoscopy, laboratory variables and vitals, and post-endoscopic results. Three machine learning models including Random Forest Classifier (RFC), Gradient Boosting Classifier (GBC), and AdaBoost Classifier were used to evaluate the need for intervention during endoscopy in our study. A total of 1372 patients were included from three major hospitals in Jordan, of whom 242 underwent endoscopic intervention. The GBC outperformed all other models and pre-endoscopic scores in predicting the need for endoscopic intervention with an area under the curve of 0.861. Findings during endoscopy, platelet count, pulse rate and systolic blood pressure during admission were the most contributing features in predicting the need for the need of endoscopic intervention. This study highlights the potential of machine learning models in enhancing decision making for UGIB management.