Background/Objectives: Acute pancreatitis (AP) is an illness that requires prompt diagnosis and treatment since it has the potential to become life-threatening. The American College of Gastroenterology 2024 (ACG24) guidelines offer a framework for diagnosis, severity, and treatment criteria. To assess Google Gemini application of ACG24 guidelines to Medical Information Mart for Intensive Care-III AP cases for risk, nutrition, and complication management. Methods: This observational cross-sectional study was based on 512 patients with AP who were treated in the Medical Information Mart for Intensive Care-III database from 2001 to 2012. The study compared the efficiency of Gemini in relation to the ACG24 guidelines in the three main areas of risk stratification, enteral nutrition timing, and necrotizing pancreatitis management. Enteral nutrition, according to the ACG24 guidelines, should be started within 48 h for patients who are capable, and antibiotics should only be used for confirmed infected necrosis. Results: The study included 512 patients who were divided into two groups: 213 patients with mild pancreatitis (41.6%) and 299 patients with severe pancreatitis (58.4%). The model achieved 85% accuracy for mild cases and 82% accuracy for severe cases of pancreatitis. The Acute Physiology and Chronic Health Evaluation II and Ranson scores matched the predictions of Gemini for both mild cases (p = 0.28 and p = 0.33, respectively) and severe cases (p = 0.31 and p = 0.27, respectively). The recommendations for early enteral nutrition and delayed feeding in mild cases were correct for 78% of patients, but the system suggested oral intake prematurely in 8% of severe cases. The antibiotic guideline compliance reached 82% among 156 patients with necrotizing pancreatitis, and the procedure for draining infected necrosis was correct 85% of the time. Conclusions: The Gemini model achieved 78–85% accuracy in determining pancreatitis severity and adherence to treatment guidelines but showed lower accuracy in nutrition timing compared to other parameters. Core Tip: This study evaluated the Google Gemini model in applying the American College of Gastroenterology 2024 guidelines for acute pancreatitis across 512 Medical Information Mart for Intensive Care-III cases. Results demonstrated 85% accuracy in severity classification, precise prediction of Acute Physiology and Chronic Health Evaluation II and Ranson scores, and 78–85% compliance with nutritional and necrotizing pancreatitis management guidelines. These findings suggest that artificial intelligence-based clinical decision support systems can provide rapid, consistent, and guideline-concordant recommendations, which are particularly valuable in settings with limited specialist expertise.

Advanced liver disease, including cirrhosis and hepatocellular carcino-ma (HCC), represents a major global health challenge, driven in part by the rising prevalence of metabolic dysfunction-associated steatotic liver disease (MASLD). In parallel, micro- and nanoplastics (MNPs) have emerged as pervasive environmental contaminants with potential hep-atotoxic effects. This review provides an in-depth analysis of current knowledge regarding the role of MNPs in the pathogenesis of cirrhosis and HCC, particularly in the context of the growing MASLD burden, and identifies key areas for future research. The literature search includ-ed original studies and review articles indexed in PubMed/MEDLINE, Web of Science, Scopus, and Google Scholar from January 1, 2014, to November 1, 2024. Evidence from animal models indicates that MNP exposure may induce hepatic changes resembling those seen in human MASLD and metabolic dysfunction-associated steatohepatitis (MASH) through both direct and indirect mechanisms. Importantly, MNPs may act as a “second hit” in the presence of pre-existing metabolic stress, potentially exacerbating liver injury. However, human data remain scarce, with only two small-scale studies investigating MNPs in clinical cohorts. Recent advances in analytical methods for quantifying MNPs in blood present new opportunities to explore their association with MASLD, cirrhosis, and HCC in human populations. While significant progress has been made in understanding MNP-induced hepatotoxicity in experimental models, their clinical relevance to human liver disease progression remains largely unexplored. Further multidisciplinary research integrating environmental science, molecular biology, and clinical hepatology is urgently needed.

Background: Colorectal cancer (CRC) remains a significant global health burden, with early detection and intervention crucial for improving patient outcomes. This study aims to develop and evaluate a novel proof-of-concept ensemble framework combining transformer-based language models and decision tree-based models for early-stage CRC screening, diagnosis, and prognosis. Methods: The ensemble framework consists of four key components: (1) GastroGPT, a transformer-based language model for extracting relevant data points from patient histories; (2) a decision tree-based model for assessing CRC risk and recommending colonoscopy; (3) GastroGPT for extracting data points from early CRC patients’ histories; and (4) a suite of decision tree-based models for predicting survival outcomes in early-stage CRC patients. The study employed a retrospective, observational, methodological design using simulated patient cases. Results: GastroGPT demonstrated high accuracy in extracting relevant data points from patient histories. The decision tree-based model for CRC risk assessment achieved an area under the receiver operating characteristic curve (AUC-ROC) of 0.85 (95% CI: 0.78–0.92) in predicting the need for colonoscopy. The decision tree-based models for survival prediction showed strong performance, with C-indices ranging from 0.71 to 0.75 for overall survival and disease-free survival at 24, 36, and 48 months. Conclusions: The novel ensemble framework demonstrates promising performance in early-stage CRC screening, diagnosis, and prognosis. Further research is needed to validate the models using larger, real-world datasets and to assess their clinical utility in prospective studies.

Objective The objective of this study is to compare guideline adherence between artificial intelligence (AI) models (Claude-3 (Anthropic, San Francisco, CA), DeepSeek-V2 (DeepSeek, Hangzhou, China), GPT-4 (OpenAI, San Francisco, CA)) and human experts in dyslipidemia management using standardized clinical scenarios based on 2019 European Society of Cardiology (ESC)/European Atherosclerosis Society (EAS) and 2021 ESC prevention guidelines. The study employed a comprehensive evaluation framework to capture the holistic nature of dyslipidemia management across multiple interconnected domains. Methods Thirty fictitious but clinically representative cases were developed by lipid specialists across five domains: cardiovascular risk assessment, lipid management, lifestyle modifications, pharmacotherapy, and special populations. This broad scope was deliberately chosen to evaluate the full complexity of integrated cardiovascular risk management as it occurs in clinical practice. Cases included all variables required for objective guideline application. AI models and clinicians (professors, specialists, residents) provided management recommendations. A blinded assessment paradigm was employed to minimize potential evaluation bias, with evaluators scoring responses using alphanumeric coding to prevent source identification bias. Responses were assessed using standardized rubrics (0-3 scales) for four equally-weighted parameters: accuracy (guideline concordance), comprehensiveness (clinical coverage), applicability (implementation feasibility), and efficacy (simulated low-density lipoprotein cholesterol (LDL-C) target attainment). Composite scores were calculated by summing all parameters (maximum 12 points). Results Correct response rates were 91% for AI, 72% for professors, 50% for specialists, and 21-32% for residents. Composite scores (mean ± SD/12) were 10.3 ± 1.0 for AI, 8.1-9.2 for professors, 7.4 ± 1.5 for specialists, and 5.2-6.2 for residents. AI excelled in literal guideline application while professors considered contextual factors (frailty, life expectancy). Professors primarily erred in LDL-C targets (using <100 vs. <55 mg/dL), while AI in nuanced risk stratification. Simulated outcomes showed LDL-C target attainment of 83% with AI, 64% with professors, and 92% with a combined approach. Conclusion AI demonstrated superior guideline adherence in standardized scenarios but may miss contextual clinical factors. The hybrid AI-human approach optimized outcomes, suggesting that augmented intelligence represents the most promising implementation strategy. Limitations include simulated cases (n = 30), potential performance bias favoring literal interpretation, and lack of real-world complexity. Prospective clinical validation is warranted.

The accurate and timely diagnosis of ileus versus volvulus is essential in emergency care, as treatment choices directly influence patient outcomes. In thi