A multicenter study demonstrates that interpretable machine learning applied to baseline OCT biomarkers accurately predicts chronicity in central serous chorioretinopathy, facilitating early risk stratification and personalized management.
This study assesses the safety, quality, and user acceptance of a conversational AI system aiding urgent primary care by engaging patients before appointments, demonstrating promising feasibility for clinical integration.
This article analyzes FDA-approved surgical AI devices, highlighting trends in technology, evidence quality, regulatory pathways, and implications for clinical adoption.
A novel deep learning model significantly improves diagnostic accuracy in distinguishing fibro-osseous jaw lesions (FD, COF, COD) from histology images, outperforming expert pathologists and offering a promising diagnostic support tool.
This review synthesizes evidence on medical interpreters’ critical roles in supporting resilience for non-English language preference patients during ambulatory diagnostic care, highlighting gaps and implications for AI interpretation
This study demonstrates promising accuracy of a patient-personalized AI system in detecting tracheostomy tube status and infant distress via video monitoring, highlighting potential for enhanced, continuous surveillance in high-risk infants
An AI-based automatic 3D virtual liver modeling tool accurately predicts intraoperative liver resection complexity (LRC) from preoperative CT scans, outperforming surgeons’ assessments and potentially improving surgical planning and p
The integration of artificial intelligence in obstetrics and gynecology necessitates clinical competence to ensure patient safety, verify outputs, and maintain physician accountability. Structured training and evaluation are essential for r
A no-code, agentic AI workflow successfully developed and externally validated a calibrated thyroid nodule malignancy classifier, demonstrating strong diagnostic performance across multiple cancer histologies and enhancing accessibility in
This study evaluates large language model (LLM) performance for predicting same-day hospital discharge, identifies primary error sources, and demonstrates that automated prompt optimization significantly improves predictive accuracy.
This study reveals that nearly 20% of online rhinoplasty photos on a major platform are digitally manipulated, using an AI model with high accuracy, highlighting the need for authenticity checks in aesthetic surgery imaging.
ChatGPT-4o achieves high accuracy on critical care board questions but shows notable limitations in image interpretation and reasoning, posing potential clinical risks.
This review synthesizes perspectives from key stakeholders on AI integration in home care, highlighting benefits, risks, and governance challenges to inform equitable, sustainable AI adoption.
A novel real-time model accurately classifies functional laryngeal behaviors from videolaryngoscopy data, enhancing clinical assessment and automation potential for voice disorder diagnostics.
This study validates the Critical View of Quality (CVQ) as an intraoperative metric of lymphadenectomy quality in distal gastrectomy and presents a novel computer vision model for its automated video-based assessment, enhancing objective su
The ACP’s 2026 position paper offers clinician-focused ethical guidance for AI use in medicine, emphasizing relationality, self-governance, and competence to safeguard patient welfare, equity, transparency, and clinical judgment.
Study found that use of an AI-driven gamified mobile app integrating home blood pressure monitoring significantly improved blood pressure control compared to usual care in hypertensive patients.
This study assesses the perceptual realism of synthetic laryngeal images created using StyleGAN3, revealing high ambiguity between real and synthetic frames and highlighting efficient training durations for realistic image generation.
A novel AI-driven risk prediction model for ovarian cancer demonstrates high accuracy and potential to improve early detection in screening settings.
A novel deep learning algorithm accurately detects moderate or greater aortic stenosis using AI-guided focused cardiac ultrasound acquired by novice operators, presenting a scalable solution to expand screening accessibility.
Open language-specific specialty feeds and department pages.