This study assessed the diagnostic accuracy and fairness of multimodal large language models (ChatGPT-4 and LLaVA) in identifying skin diseases across various demographic groups. Analysis of approximately 10,000 medical images showed that while these AI models generally outperform traditional approaches, biases in performance related to sex and age were evident, particularly with LLaVA showing clear sex-related disparities.

Researchers advocate for attention to demographic fairness in AI-driven healthcare solutions. Further studies are planned to include additional demographic factors such as skin tone, aiming to enhance AI usability and reliability across diverse patient populations.

Until today, skin, brain, and all tissues of the human body were difficult to observe in detail with an optical microscope, since the contrast in the image was hindered by the high density of their structures. The research group of the Molecular Microscopy and Spectroscopy Lab at the Istituto Italiano di Tecnologia (IIT-Italian Institute of Technology) in Genoa has devised a new method that allows scientists to see and photograph biological samples in all their complexity, obtaining clear and detailed images. The new technique has been made available to the scientific community in “open science” mode, representing an advantage in the biomedical field, since it allows us to observe active cells, even in the presence of diseases, as well as to understand how drugs interact with living tissues.

BASILISK, the global esports organization built to champion science, debuted historic partnerships with the Institute for Quantum Information and Matter (IQIM) at the California Institute of Technology and The Planetary Society at the Esports World Cup 2025 in Riyadh, Saudi Arabia. These partnerships create a powerful bridge between renowned scientific institutions and the next generation of STEM enthusiasts. Together, they amplify each other’s missions: expanding the reach of science advocacy to millions of science enthusiasts at a critical moment for the field.

A research team developed a novel method using bidirectional reflectance factor (BRF) spectra combined with the PROSPECT-PRO model and modified ratio indices to estimate nitrogen content nondestructively.

Xiaowei Teng, James H. Manning Professor of Chemical Engineering, has been awarded a $395,886 grant from the National Science Foundation to study an electrochemical system that could be used to address one of the major challenges in water treatment: the buildup of insoluble mineral deposits. 

A new method harnessing the Cyclone Global Navigation Satellite System (CYGNSS) satellite constellation significantly improves the accuracy of soil moisture retrieval by eliminating vegetation interference and removing dependency on external datasets.