Regulation of Artificial Intelligence in Health Care and Biomedicine

JAMA, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 20, 2025

Language: Английский

---

Regulation of Artificial Intelligence in Health Care and Biomedicine—Reply

JAMA, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 20, 2025

Language: Английский

---

A systematic review of machine learning-based prognostic models for acute pancreatitis: Towards improving methods and reporting quality

PLoS Medicine, Journal Year: 2025, Volume and Issue: 22(2), P. e1004432 - e1004432

Published: Feb. 24, 2025

Background An accurate prognostic tool is essential to aid clinical decision-making (e.g., patient triage) and advance personalized medicine. However, such a lacking for acute pancreatitis (AP). Increasingly machine learning (ML) techniques are being used develop high-performing models in AP. methodologic reporting quality has received little attention . High-quality study methodology critical model validity, reproducibility, implementation. In collaboration with content experts ML methodology, we performed systematic review critically appraising the of recently published AP models. Methods/findings Using validated search strategy, identified studies from databases MEDLINE EMBASE between January 2021 December 2023. We also searched pre-print servers medRxiv, bioRxiv, arXiv pre-prints registered Eligibility criteria included all retrospective or prospective that developed new existing patients predicted an outcome following episode Meta-analysis was considered if there homogeneity design type predicted. For risk bias (ROB) assessment, Prediction Model Risk Bias Assessment Tool. Quality assessed using Transparent Reporting Multivariable Individual Prognosis Diagnosis—Artificial Intelligence (TRIPOD+AI) statement defines standards 27 items should be reported publications The strategy 6,480 which 30 met eligibility criteria. Studies originated China (22), United States (4), other (4). All none sought validate model, producing total 39 severity (23/39) mortality (6/39) were most common outcomes mean area under curve endpoints 0.91 (SD 0.08). ROB high at least one domain models, particularly analysis (37/39 models). Steps not taken minimize over-optimistic performance 27/39 Due heterogeneity how defined determined, meta-analysis performed. on only 15/27 TRIPOD+AI standards, 7/30 justifying sample size 13/30 assessing data quality. Other deficiencies omissions regarding human–AI interaction (28/30), handling low-quality incomplete practice (27/30), sharing analytical codes (25/30), protocols source (19/30). Conclusions There significant based patients. These undermine implementation these despite their promise superior predictive accuracy. Registration Research Registry (reviewregistry1727)

Language: Английский

---

Augmenting Insufficiently Accruing Oncology Clinical Trials Using Generative Models: Validation Study

Journal of Medical Internet Research, Journal Year: 2025, Volume and Issue: 27, P. e66821 - e66821

Published: March 5, 2025

Insufficient patient accrual is a major challenge in clinical trials and can result underpowered studies, as well exposing study participants to toxicity additional costs, with limited scientific benefit. Real-world data provide external controls, but insufficient affects all arms of study, not just controls. Studies that used generative models simulate more patients were the scenarios considered, replicability criteria, number models, evaluated. This aimed perform comprehensive evaluation on extent be compensate for trials. We performed retrospective analysis using 10 datasets from 9 fully accrued, completed, published cancer For each trial, we removed latest recruited (from 10% 50%), trained model remaining patients, simulated replace ones augment available data. then replicated this augmented dataset determine if findings remained same. Four different evaluated: sequential synthesis decision trees, Bayesian network, adversarial variational autoencoder. These compared sampling replacement (ie, bootstrap) simple alternative. Replication analyses 4 metrics: agreement, estimate standardized difference, CI overlap. Sequential replication metrics removal up 40% last (decision agreement: 88% 100% across datasets, 100%, cannot reject difference null hypothesis: overlap: 0.8-0.92). Sampling was next most effective approach, agreement varying 78% 89% datasets. There no evidence monotonic relationship estimated effect size recruitment order these studies. suggests earlier trial systematically than those later, at least partially explaining why early effectively later trial. The fidelity generated relative training Hellinger distance high cases. an oncology few 60% target recruitment, enable simulation full had continued accruing alternative drawing conclusions study. results demonstrating potential rescue poorly trials, studies are needed confirm generalize them other diseases.

Language: Английский

---

The Value of Clinical Decision Support in Healthcare: A Focus on Screening and Early Detection

Diagnostics, Journal Year: 2025, Volume and Issue: 15(5), P. 648 - 648

Published: March 6, 2025

In a rapidly changing technology landscape, “Clinical Decision Support” (CDS) has become an important tool to improve patient management. CDS systems offer medical professionals new insights diagnostic accuracy, therapy planning, and personalized treatment. addition, provide cost-effective options augment conventional screening for secondary prevention. This review aims (i) describe the purpose mechanisms of systems, (ii) discuss different entities algorithms, (iii) highlight quality features, (iv) challenges limitations in clinical practice. Furthermore, we (v) contemporary algorithms oncology, acute care, cardiology, nephrology. particular, consolidate research on across diseases that imply significant disease economic burden, such as lung cancer, colorectal hepatocellular coronary artery disease, traumatic brain injury, sepsis, chronic kidney disease.

Language: Английский

---

Artificial Intelligence-Based Methods for Drug Repurposing and Development in Cancer

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(5), P. 2798 - 2798

Published: March 5, 2025

Drug discovery and development remains a complex time-consuming process, often hindered by high costs low success rates. In the big data era, artificial intelligence (AI) has emerged as promising tool to accelerate optimize these processes, particularly in field of oncology. This review explores application AI-based methods for drug repurposing natural product-inspired design cancer, focusing on their potential address challenges limitations traditional approaches. We delve into various approaches (machine learning, deep others) that are currently being employed purposes, role experimental techniques By systematically reviewing literature, we aim provide comprehensive overview current state AI-assisted cancer workflows, highlighting AI’s contributions accelerating development, reducing costs, improving therapeutic outcomes. also discusses opportunities associated with integration AI pipeline, such quality, interpretability, ethical considerations.

Language: Английский

---

Consternation as Congress proposal for autonomous prescribing AI coincides with the haphazard cuts at the FDA

npj Digital Medicine, Journal Year: 2025, Volume and Issue: 8(1)

Published: March 18, 2025

We live in interesting regulatory times. In January, a bill was introduced to the US Congress proposing that AI "can qualify as practitioner eligible prescribe drugs" if overseen by States and FDA. This bold contentious move. Even proponents of AI's swift integration into medicine must recognize deep paradox: this proposal emerges even FDA's world-leading infrastructure for oversight faces dismantling.

Language: Английский

---

Computational pathology: A comprehensive review of recent developments in digital and intelligent pathology

Published: March 1, 2025

Language: Английский

---

AI-guided precision parenteral nutrition for neonatal intensive care units

Nature Medicine, Journal Year: 2025, Volume and Issue: unknown

Published: March 25, 2025

One in ten neonates are admitted to neonatal intensive care units, highlighting the need for precise interventions. However, application of artificial intelligence (AI) guiding remains underexplored. Total parenteral nutrition (TPN) is a life-saving treatment preterm neonates; however, implementation therapy its current form subjective, error-prone and resource-consuming. Here, we developed TPN2.0—a data-driven approach that optimizes standardizes TPN using information collected routinely electronic health records. We assembled decade compositions (79,790 orders; 5,913 patients) at Stanford train TPN2.0. In addition internal validation, also validated our model an external cohort (63,273 3,417 from second hospital. Our algorithm identified 15 formulas can enable precision-medicine (Pearson's R = 0.94 compared experts), increasing safety potentially reducing cost. A blinded study (n 192) revealed physicians rated TPN2.0 higher than best practice. patients with high disagreement between actual prescriptions TPN2.0, standard were associated increased morbidities (for example, odds ratio 3.33; P value 0.0007 necrotizing enterocolitis), while recommendations linked reduced risk. Finally, demonstrated employing transformer architecture enabled guideline-adhering, physician-in-the-loop allow collaboration team AI. An defines set total assist clinicians personalized able adapt patient status, validation large cohorts reader study.

Language: Английский

---

Integrating Artificial Intelligence in Youth Mental Health Care: Advances, Challenges, and Future Directions

Current Treatment Options in Psychiatry, Journal Year: 2025, Volume and Issue: 12(1)

Published: April 10, 2025

Language: Английский

---