Citations

Background Large language models (LLMs) are entering clinical workflows, yet their alignment with medical-ethical principles is unclear. Objective To evaluate how LLMs reconcile deontological and utilitarian reasoning in medically adapted “Trolley dilemma” scenarios. Methods We tested 14 LLMs, including GPT-o1-preview and DeepSeek-R1-Distill-Llama, using five medically adapted Trolley dilemmas sourced from philosophical and real-world scenarios. Each model generated 10 independent answers per case (overall 700 responses). Responses were forced to “yes” (utilitarian) or “no” (deontological). The primary outcome was the proportion of utilitarian choices. χ 2 tests compared models. Results LLM responses varied. Some models, such as GPT-4 and Gemma, selected the utilitarian option in 20% of cases, while others, like Qwen-2-7B, reached 80% ( p < .001). GPT-o1-preview and DeepSeek-R1-Distill-Llama selected utilitarian responses in 44% and 38% of cases, respectively. Some models selected the consequence-maximizing option in vignettes where doing so required overriding consent or endorsing intentional harm within the scenario framing. Seven of the fourteen models tested recommended impermissible medical actions, including nonconsensual limb amputation, killing a healthy person for organ harvesting, and transferring a patient to hospice against their wishes, in 44 of 700 outputs (6.3 %). Conclusions In this constrained forced-choice stress test, models occasionally endorsed boundary-violating actions within the vignette framing. These results do not generalize to routine clinical decision-making, but motivate further benchmarked evaluation of refusal behavior and boundary constraints in ethically sensitive medical contexts. We must carefully define how these models influence clinical decisions, establish robust ethical guidelines for their use, and ensure that fundamental ethical norms are not compromised.

ABSTRACT Artificial intelligence (AI), encompassing machine learning, deep learning and generative AI, is poised to redefine nephrology by enabling earlier detection, more precise risk stratification and workflow-integrated clinical decision support across the spectrum of kidney disease. This state-of-the-art review synthesizes emerging applications of AI in acute kidney injury (AKI), chronic kidney disease (CKD), dialysis and kidney transplantation, with attention to clinical integration, real-world validation, workflow implementation and translational challenges. In AKI, predictive models trained on high-frequency electronic health record data and intensive care unit telemetry have demonstrated strong performance in forecasting critical events, yet translation into routine clinical workflows remains limited. In CKD, machine learning tools support progression risk stratification and phenotype clustering, with the potential to inform individualized surveillance and therapy. AI-enabled dialysis management systems optimize ultrafiltration, anemia control and vascular access surveillance, while generative AI and large language models offer novel capabilities for clinical documentation, triage and patient education. In transplantation, AI applications span organ allocation, dynamic graft monitoring and digital pathology-assisted rejection classification, with validated tools such as the iBox system gaining regulatory recognition. Implementation challenges include data heterogeneity, bias, interpretability, regulatory uncertainty and workflow integration. Looking ahead, multimodal integration of imaging, pathology and multi-omics data may support biologically informed precision nephrology. Reinforcement learning, digital twins and ambient intelligence are emerging as adaptive decision-support paradigms rather than autonomous care systems. Regulatory frameworks are evolving to accommodate adaptive algorithms, underscoring the need for clinician engagement in model development, validation and deployment. As AI matures from pilot to practice, nephrologists who embrace and help shape these tools will lead the transition toward a more personalized, efficient and equitable future for kidney care.

Background Artificial intelligence (AI)-generated text to video is emerging in medical education, but its effectiveness, accuracy and safety remain uncertain. We aimed to synthesise empirical studies evaluating these tools in learner or patient education. Methods A comprehensive search was conducted in MEDLINE/PubMed, Google Scholar, Scopus, Cochrane Review and Web of Science for studies published up to January 2025. Eligible studies evaluated AI-generated text to video for medical or patient education, reporting both quantitative and qualitative outcomes. Two reviewers screened and extracted data. The review adhered to Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Results Out of 103 identified studies, five studies met the inclusion criteria: four evaluated patient education and one evaluated physician training. Clinical areas were ophthalmology (2/5), plastic surgery (1/5), dysphagia rehabilitation (1/5) and neurosurgical training (1/5). In ophthalmology, control materials outperformed AI-generated videos on image/script accuracy (p<0.005), with similar script-image alignment. In dysphagia rehabilitation, a randomised trial reported improvements in swallowing function and related outcomes with an AI-assisted video game intervention (p<0.001). A plastic surgery study reported greater user preference for a video avatar tool compared with a text chatbot (63.5% vs 28.1%). Across the reviewed studies, samples were small and CIs were rarely reported. Outcome measures were heterogeneous, and meta-analysis was not feasible. Conclusion AI-generated videos can enhance engagement or selected outcomes in certain contexts, yet concerns about accuracy and inconsistent measurement persist. Current evidence is sparse and mixed. Currently, these tools can complement, rather than replace, standard resources until non-inferiority is demonstrated for the prespecified primary outcomes. PROSPERO registration number CRD42025640042.

Large language models (LLMs) have the potential to transform healthcare by assisting in documentation, diagnosis, patient communication, and medical education. However, their integration into clinical practice remains a challenge. This perspective explores the barriers to implementation by synthesizing recent evidence across five challenge domains: workflow misalignment and diagnostic safety, bias and equity, regulatory and legal governance, technical vulnerabilities such as hallucinations or data poisoning, and the preservation of patient trust and human connection. While the perspective focuses on barriers, LLM capabilities and mitigation strategies are advancing rapidly, raising the likelihood of near-term clinical impact. Drawing on recent empirical studies, we propose a framework for understanding the key technical, ethical, and practical challenges associated with deploying LLMs in clinical environments and provide directions for future research, governance, and responsible deployment.

Multimodal technology is poised to revolutionize clinical practice by integrating artificial intelligence with traditional diagnostic modalities. This evolution traces its roots from Hippocrates’ humoral theory to the use of sophisticated AI-driven platforms that synthesize data across multiple sensory channels. The interplay between historical medical practices and modern technology challenges conventional patient–clinician interactions and redefines diagnostic accuracy. Highlighting applications from neurology to radiology, the potential of multimodal technology emerges, suggesting a future where AI not only supports but enhances human sensory inputs in medical diagnostics. This shift invites the medical community to navigate the ethical, practical, and technological changes reshaping the landscape of clinical medicine.

Abstract Objectives Social support (SS) and social isolation (SI) are social determinants of health (SDOH) associated with psychiatric outcomes. In electronic health records (EHRs), individual-level SS/SI is typically documented in narrative clinical notes rather than as structured coded data. Natural language processing (NLP) algorithms can automate the otherwise labor-intensive process of extraction of such information. Materials and Methods Psychiatric encounter notes from Mount Sinai Health System (MSHS, n = 300) and Weill Cornell Medicine (WCM, n = 225) were annotated to create a gold-standard corpus. A rule-based system (RBS) involving lexicons and a large language model (LLM) using FLAN-T5-XL were developed to identify mentions of SS and SI and their subcategories (eg, social network, instrumental support, and loneliness). Results For extracting SS/SI, the RBS obtained higher macroaveraged F1-scores than the LLM at both MSHS (0.89 versus 0.65) and WCM (0.85 versus 0.82). For extracting the subcategories, the RBS also outperformed the LLM at both MSHS (0.90 versus 0.62) and WCM (0.82 versus 0.81). Discussion and Conclusion Unexpectedly, the RBS outperformed the LLMs across all metrics. An intensive review demonstrates that this finding is due to the divergent approach taken by the RBS and LLM. The RBS was designed and refined to follow the same specific rules as the gold-standard annotations. Conversely, the LLM was more inclusive with categorization and conformed to common English-language understanding. Both approaches offer advantages, although additional replication studies are warranted.

AbstractLarge language models (LLMs) are transforming the field of natural language processing (NLP). These models offer opportunities for radiologists to make a meaningful impact in their field. NLP is a part of artificial intelligence (AI) that uses computer algorithms to study and understand text data. Recent advances in NLP include the Attention mechanism and the Transformer architecture. Transformer-based LLMs, such as GPT-4 and Gemini, are trained on massive amounts of data and generate human-like text. They are ideal for analysing large text data in academic research and clinical practice in radiology. Despite their promise, LLMs have limitations, including their dependency on the diversity and quality of their training data and the potential for false outputs. Albeit these limitations, the use of LLMs in radiology holds promise and is gaining momentum. By embracing the potential of LLMs, radiologists can gain valuable insights and improve the efficiency of their work. This can ultimately lead to improved patient care.

Background: Venous thromboembolism (VTE) is a life-threatening vascular event with environmental and genetic determinants. Recent VTE genome-wide association studies (GWAS) meta-analyses involved nearly 30 000 VTE cases and identified up to 40 genetic loci associated with VTE risk, including loci not previously suspected to play a role in hemostasis. The aim of our research was to expand discovery of new genetic loci associated with VTE by using cross-ancestry genomic resources. Methods: We present new cross-ancestry meta-analyzed GWAS results involving up to 81 669 VTE cases from 30 studies, with replication of novel loci in independent populations and loci characterization through in silico genomic interrogations. Results: In our genetic discovery effort that included 55 330 participants with VTE (47 822 European, 6320 African, and 1188 Hispanic ancestry), we identified 48 novel associations, of which 34 were replicated after correction for multiple testing. In our combined discovery-replication analysis (81 669 VTE participants) and ancestry-stratified meta-analyses (European, African, and Hispanic), we identified another 44 novel associations, which are new candidate VTE-associated loci requiring replication. In total, across all GWAS meta-analyses, we identified 135 independent genomic loci significantly associated with VTE risk. A genetic risk score of the significantly associated loci in Europeans identified a 6-fold increase in risk for those in the top 1% of scores compared with those with average scores. We also identified 31 novel transcript associations in transcriptome-wide association studies and 8 novel candidate genes with protein quantitative-trait locus Mendelian randomization analyses. In silico interrogations of hemostasis and hematology traits and a large phenome-wide association analysis of the 135 GWAS loci provided insights to biological pathways contributing to VTE, with some loci contributing to VTE through well-characterized coagulation pathways and others providing new data on the role of hematology traits, particularly platelet function. Many of the replicated loci are outside of known or currently hypothesized pathways to thrombosis. Conclusions: Our cross-ancestry GWAS meta-analyses identified new loci associated with VTE. These findings highlight new pathways to thrombosis and provide novel molecules that may be useful in the development of improved antithrombosis treatments.