Ranks distinct AI gain and problem claims from the published record. Scores reward impact, independent source strength, scale, confidence, and recency.
Researchers built and tested machine learning models to predict cardiac immune-related adverse events shortly after starting immune checkpoint inhibitor therapy. Using records for 61,117 patients from TriNetX from 2010 to 2023, they defined events by diagnosis codes within 90 days plus hospital visits and compared elastic net logistic regression, gradient boosted trees, and random forest models.
Researchers developed metabolomics-based machine learning models to stratify future major depressive disorder risk among 41,459 obese participants followed for a median of 14.4 years. The optimized LightGBM model achieved AUCs of 0.844, 0.824 and 0.834 for 3-, 5- and 9-year predictions and outperformed existing clinical models, with temporal validation showing AUCs of 0.738-0.776.
This opinion article examines the concept of clinical justification for osteoporosis imaging, noting that advances in imaging and AI now permit opportunistic identification of osteoporosis and other conditions beyond the original scan purpose, and discusses how these opportunistic tools might eventually become standalone justified diagnostic pathways.
Researchers audited 5,000 archived liquid-based cytology cases and compared an interpretable machine learning pipeline, LBC-NET, to a double-blind panel of senior cytopathologists for detection of high-grade squamous intraepithelial lesions or worse, testing robustness across age and clinical subgroups.
By July 2026, researchers had surveyed 69 Japanese university students about everyday GenAI use and analyzed responses with systematic grounded theory. They found students moved through an iterative moral trajectory from recognizing power and risk to feeling ethical anxiety, then to reflexive evaluation and conditional trust, captured in the NECoAI model.
A 2026 position paper examines AI systems that record voice and video during pediatric emergencies, noting they are emerging as HCI technologies with implications for clinical work and are promoted for documentation, team performance, and debriefing. The authors argue that clinicians, parents, and child patients have been largely absent from design and governance.
What happened is that by 2026-06-03 scholars observed that smart technology, AI, robotics and algorithms were changing work design, with reviews noting varied effects on performance and wellbeing and primary emphasis on displacement of routine tasks and the need to upskill and reskill workers.
By July 2026, researchers examined singing data collection as AI voice synthesis advanced, analyzing three singing datasets with the Ethically Aligned Stakeholder Elicitation framework. They found data-contributors have event-centric roles with minimal authority over licensing and access, while data-collectors retain control.
ace