AI-enabled medical devices promise improved medical care and support for healthcare professionals. However, the safety and performance of such systems not only depends on algorithms or technical specifications. It is equally important how people use these devices and applications. In a recent publication in the scientific journal NEJM AI, a research team led by Prof. Stephen Gilbert from Else Kröner Fresenius Center (EKFZ) for Digital Health at TUD Dresden University of Technology systematically analyzes risks that can arise in human-AI interactions and makes recommendations for manufacturers and regulatory evaluators.

How does the human brain’s electrical activity grow from childhood, peak in adulthood, and decline in older age? A multinational team has tackled this question by linking the brain’s “wiring diagram” and signal‑conduction speed to two familiar features of an electroencephalogram (EEG): the broadband background activity (ξ, pronounced “xi”) and the more rhythmic alpha waves. Their work, published in National Science Review, introduces a new model called Xi–αNET (“Xi–AlphaNET”) that explains how anatomical connections and nerve‑signal delays give rise to these patterns and how they change over the lifespan.

Sepsis complicated by acute respiratory failure is a severe condition among patients admitted to the intensive care unit (ICU). Researchers have developed and externally validated a machine learning model to predict the 28-day mortality risk in such ICU patients with sepsis. Using routinely collected clinical variables from the first 24 hours after admission, the model demonstrated stable predictive performance in large critical care databases, with potential value for early risk stratification and individualized treatment decision-making.

By designing a hybrid system with variable-sized neurons, the key problems in the manufacturing process of ODNNs were successfully resolved, thereby achieving high performance and high reliability in the practical application of ODNNs.

This article provides a comprehensive review of the research progress in perovskite light-emitting diodes (PeLEDs) up to the end of 2025, systematically compiling their key performance parameters (including EQE, luminance, and lifetime) and deeply discussing the main challenges and future directions for commercialization, such as blue efficiency, active-matrix display integration, and lead-free development.