4252 papers
This study presents GB-DualmodalNet, a clinical-prompt-enhanced deep learning framework that integrates preoperative ultrasound images with natural language-processed clinical data to accurately predict gallbladder polypoid lesion subtypes, thereby improving diagnostic performance over image-only models and reducing unnecessary cholecystectomies.
This study developed and validated a simple, seven-point risk scoring scale based on five independent factors (age, lifting/carrying ability, maximum walking distance, finger-ring test, and mood status) to effectively screen for sarcopenia among community-dwelling older adults in China using the 2025 AWGS criteria, demonstrating good discrimination, calibration, and clinical utility for primary care settings.
This study demonstrates that machine learning models, particularly TabICL and TabPFN, can effectively predict weaning trajectories and extubation success in mechanically ventilated patients early in the weaning process by leveraging clinically meaningful and potentially modifiable factors such as organ dysfunction, frailty, and sedation state.
This exploratory convergent mixed-methods study demonstrates that a zero-cost, first-person creative narrative activity ("When Drugs Speak") significantly enhances undergraduate medical students' engagement, retention, and conceptual understanding of anti-seizure medications by leveraging retrieval practice and elaborative encoding, as evidenced by high student preference and qualitative themes of narrative schema-building.
This retrospective study of 59 patients demonstrates that individualized multilevel swallowing improvement surgeries significantly enhance swallowing function and oral intake in those with neurogenic dysphagia, despite a notable postoperative complication rate.
This single-center retrospective study of 618 cirrhosis patients found that a preprocedural eGFR below 60 mL/min/1.73 m² is associated with increased all-cause mortality after covered-stent TIPS, though this association is attenuated when adjusting for MELD scores, suggesting eGFR is a useful but non-independent marker for pre-TIPS risk assessment.
This study developed and validated a combined risk factor prediction model using data from the Linxian Nutrition Intervention Trial cohort, which effectively stratifies individual risk for esophageal squamous cell carcinoma based on eight independent factors to facilitate targeted early screening and prevention.
In a randomized controlled trial involving 72 children aged 8–9, a six-week school-based high-intensity interval training (HIIT) program was found to improve objectively measured classroom engagement more effectively than moderate-intensity continuous training (MICT) or no exercise.
This study demonstrates that fibroblast activation protein (FAP) is upregulated in injured arteries following peripheral arterial disease surgery, where it promotes vascular restenosis by driving smooth muscle cell migration via the Net1–RhoA–ROCK signaling axis, highlighting its potential as a therapeutic target.
In a propensity score-matched retrospective cohort study of 1,238 patients undergoing laparoscopic hysterectomy, lactulose enema bowel preparation was associated with a significantly higher likelihood of early postoperative C-reactive protein elevation compared to mechanical enema, suggesting that bowel preparation regimens should be evaluated as specific exposures rather than a binary variable.