Following this puts new work involving it at the top of your briefing, with a note saying why it is there. Links are taken from the source record, never inferred.
A diagnostic benchmark study introducing a new evaluation framework for VLM behavior under uncertainty; reports descriptive findings and model comparisons without randomization, control, or clinical outcome validation, appropriate for a preprint stress-testing methodology.
Early-phase preclinical study demonstrating a novel imaging probe in mice with exploratory findings on diet-enzyme relationships; lacks human validation, clinical endpoints, and comparative controls.
Retrospective cohort study with clear stratification and identified prognostic factors in radioiodine-refractory DTC, but limited by single-centre design, modest sample size, and surrogate endpoint assessment in the non-refractory group.
Retrospective cohort with small sample (n=44), no comparator arm, and surrogate endpoints; findings are descriptive of age-related differences but lack statistical power and prospective design to guide clinical practice.