Artificial intelligence (AI) is increasingly being explored as a tool to support clinical decision-making, yet its real-world performance in pediatric diagnosis remains unclear. Now, a Pediatric Investigation study using authentic clinical cases reports that advanced AI models outperform clinicians in diagnostic accuracy, particularly for rare diseases, while a combined human-AI approach achieves the highest overall success. The findings highlight the potential of AI as a complementary tool to improve diagnostic precision and patient outcomes.

A new study shows that systems designed to capture methane from cow manure, called dairy digesters, are highly effective. But on the rare occasions they fail, the leaks are large enough to offset their climate benefits.

Researchers at The University of Manchester have created a groundbreaking physics‑informed machine‑learning model that can run molecular simulations for unprecedented lengths of time, even at temperatures as high as 1000 Kelvin.

Both Pd single atoms (Pd₁) and clusters (Pd_c) were constructed in three-dimensional ordered macroporous (3DOM) In₂O₃ for photocatalytic CO₂ reduction with H₂O. The large surface area and abundant pore channels of 3DOM-In₂O₃ facilitate mass transfer and intermediate enrichment. The synergisticPd₁ and Pd_c active sites enhance the adsorption and activation of CO₂ and H₂O. The localized surface plasmon resonance of Pd clusters induces a photothermal effect, further accelerating the reaction kinetics.

Researchers at the Agricultural Genomics Institute at Shenzhen, Chinese Academy of Agricultural Sciences (CAAS), have engineered a high-fidelity adenine base editor variant, ABE8eY149V, to overcome the severe genome-wide off-target risks associated with the hyperactive ABE8e. Through comprehensive genome-wide off-target analysis by two-cell embryo injection (GOTI), the team first revealed that ABE8e induces massive sgRNA-independent single-nucleotide variants (SNVs), nearly 30-fold higher than spontaneous background levels. 

Focusing on the engineering challenge of achieving stable, high-strength welding between rough metals surfaces and transparent materials, this work provides an in-depth elucidation of the femtosecond laser welding mechanism for dissimilar materials under non-optical-contact conditions. Through high-speed in situ imaging techniques, it reveals the dynamic coupling between linear absorption in the metal and nonlinear absorption in sapphire during ultrafast laser irradiation. The study further identifies an active interfacial gap filling effect of molten metal, which proactively regulates the free space region at the interface. It clarifies that the welding strength is primarily limited by cracks induced by thermal stress in sapphire, and demonstrates welding performance exceeding 10 MPa between rough Invar alloy and sapphire. These findings offer theoretical guidance and technical support for high-strength, highly stable welding of dissimilar materials.