A new study suggests that a soy-derived lysolecithin, LPC70, reduced high-salt diet (HSD)-induced hypertension, social behavior deficits, and object recognition memory impairment in mice. The benefits are associated with receptor-specific regulation in prostaglandin (PG) signaling. An increased production of arachidonic acid-derived PGs was also noted. The findings suggest LPC70 may serve as a functional food ingredient to mitigate HSD-associated cardiovascular and cognitive decline.

Keeping hydrogen liquid at about 20 K demands exceptional insulation, since small heat leaks generate boil-off gas and raise tank pressure. A new study used 3D multiphase simulations to evaluate a semi-active system in which a secondary cryogenic coolant, liquefied natural gas or liquid nitrogen, circulates through embedded insulation channels, intercepting heat before it reaches the inner tank. In optimized configurations, simulated heat ingress fell by 43.6% with both coolants, guiding cooling-channel placement and design.

This review systematically bridges non-resonant and resonant metasurfaces, highlighting their complementary roles in shaping next-generation photonic devices. It covers physical mechanisms, key structures, and transformative applications in imaging, color routing, and displays. The fusion of both paradigms promises multifunctional, chip‑integrated meta‑devices.

This review synthesizes collective insights from the 1st China–Australia–New Zealand Water Quality Modeling Symposium (CANWaS 2024) and outlines a roadmap for strengthening international collaboration in surface water quality modeling. Organized by Jiping Jiang from Harbin Institute of Technology and Danlu Guo from Australian National University, an international research team identifies shared priorities, mutual interests, and actionable pathways—including best-practice guidelines, open knowledge sharing, and joint technological innovation—to advance predictive capacity and policy impact for global freshwater management.

A new Perspective in Science Bulletin explains why computational pathology must move beyond accurate algorithms. To help in real hospitals, AI tools need to be interpretable, reportable, cost-aware, and tested in routine clinical workflows.