New AI tool detects hidden warning signs of disease
Peer-Reviewed Publication
This month, we’re focusing on artificial intelligence (AI), a topic that continues to capture attention everywhere. Here, you’ll find the latest research news, insights, and discoveries shaping how AI is being developed and used across the world.
Updates every hour. Last Updated: 7-Jan-2026 18:11 ET (7-Jan-2026 23:11 GMT/UTC)
McGill University researchers have developed an artificial intelligence tool that can detect previously invisible disease markers inside single cells.
In a study published in Nature Communications, the researchers demonstrate how the tool, called DOLPHIN, could one day be used by doctors to catch diseases earlier and guide treatment options.
On October 7, 2025, Karlsruhe Institute of Technology (KIT) celebrates its anniversary. On the same day, exactly 200 ago, Ludwig I, Grand Duke of Baden, signed the founding decree for the Karlsruhe Polytechnic School, the first predecessor institution of today’s University of Excellence. The Grand Duke founded it out of his “concern for the education of our dear and loyal bourgeoisie.” Today, KIT is a place where science is shaping the future by contributing energy, mobility, and climate research as well as robotics and artificial intelligence.
ChatGPT has not decreased activity on the world’s largest online encyclopaedia, but AI data scrapers and the influence of Large Language Models still cast a shadow over its future research suggests.
New research published in Nature Ecology & Evolution sheds light on the timelines and pathways of evolution of fungi, finding evidence of their influence on ancient terrestrial ecosystems. The study, led by researchers from the Okinawa Institute of Science and Technology (OIST) and collaborators, indicates the diversification of fungi hundreds of millions of years before the emergence of land plants.
Dr Bianca Moffett at the SAMRC/Wits-Agincourt Unit leads the AfriCAT project, which entails building a first-of-its-kind adaptive testing tool to inform measurement-based mental healthcare for depression and anxiety among adolescents in Africa.
The AfriCAT tool is based on Computerised Adaptive Testing, a novel approach to mental health assessment. Unlike most traditional assessments, which ask a standard set of questions to all users, Computerised Adaptive Tests are based on advanced statistical and machine learning methods, which use a person’s initial responses to select the next best questions. The goal of adaptive testing is to use as few questions as possible while still making an exact assessment, tailored to the individual.
A recent study from the University of Eastern Finland found that poorer performance in mathematics predicts greater mathematics anxiety. The study also suggests that, on average, girls experience more mathematics anxiety than boys.
The University of Osaka researchers developed a reinforcement learning framework that lets dialogue systems efficiently learn new words while asking fewer questions. This approach enhances user experience and enables future systems to naturally acquire family-specific nicknames and expressions, becoming more familiar companions in everyday life.