4376 papers
This study identifies age, acute cerebral infarction, and hydrocephalus as independent predictors of poor long-term prognosis in intracranial tuberculosis patients, demonstrating that contrast-enhanced MRI is a critical tool for early risk assessment and clinical management.
This pilot study introduces and validates the 3-Dimensional Accuracy Score (3DAS), a new composite metric for evaluating the anatomical fidelity of 3D-printed models and STL files, revealing that segmentation quality during DICOM-to-STL conversion is the primary source of error and highlighting the need for standardized verification in outsourced dental workflows.
This retrospective single-center study of 769 ICU admissions demonstrates that random forest models achieve superior mortality prediction (AUC 0.91) compared to other machine learning approaches, while identifying surgical status, sepsis, and acute severity markers as key clinically informative predictors.
This retrospective cohort study of 5,727 preterm infants in China demonstrates that postmenstrual age at first retinopathy of prematurity diagnosis is an independent prognostic marker for disease severity and retreatment burden, with the extremely low gestational age (<25 weeks) group facing significantly higher risks and a disproportionate need for multiple interventions, thereby supporting the adoption of gestational age-stratified screening schedules.
This study developed and externally validated an interpretable machine-learning model using only clinical variables available at diagnosis to accurately predict synchronous distant metastasis in invasive bladder cancer, demonstrating superior performance over traditional TNM staging and offering a freely accessible online tool for individualized pre-treatment risk assessment.
This study demonstrates that ATR-FTIR spectroscopy combined with machine learning analysis of blood serum can effectively distinguish colorectal cancer patients from healthy controls with high accuracy (90.9%), sensitivity (88.9%), and specificity (92.9%), offering a promising minimally invasive diagnostic tool.
This study developed and validated the 18-item Clinical Contraceptive Knowledge Test (CCKT-18) using Item Response Theory among Vietnamese medical students, demonstrating it to be a robust, unidimensional, and precise instrument for assessing contraceptive knowledge and guiding curriculum improvements.
This retrospective multicenter study demonstrates that time to metastasis (TTM) is an independent prognostic factor in metastatic urothelial carcinoma and that incorporating it into the Bellmunt score significantly improves risk stratification accuracy and clinical utility.
This retrospective cohort study of over 77,000 Belgian adults demonstrates that inhaler device type is independently associated with therapeutic adherence, with dry powder inhalers (DPIs) showing significantly higher adherence rates compared to pressurised metered-dose inhalers (pMDIs) and soft mist inhalers (SMIs) across various pharmacological classes.
This systematic review of randomized controlled trials in high-income countries found no strong evidence that artificial intelligence-assisted interventions effectively reduce alcohol consumption or smoking, largely due to mixed results, the inability to isolate AI's specific impact within multi-component interventions, and a high risk of bias across the included studies.