Researchers at Osaka Metropolitan University developed models that classify X-ray images into specific body regions and simultaneously determine the imaging method and image orientation. Using these models, they successfully classified almost all data for use in deep-learning models.

Gastric (stomach) cancer remains one of the most common and deadly cancers in East Asia, including Korea. Yet despite its high prevalence, it has received far less molecular attention than colorectal cancer, which is more common in Western countries. As a result, many of today’s models of gastric cancer biology are still based on assumptions borrowed from colorectal cancer research — often with limited success when applied to patients.

One of the biggest unanswered questions has concerned the very first steps of gastric cancer development: how do early cancer cells survive and grow when they should not?

Under normal conditions, cells lining the stomach cannot grow independently. They rely on constant signals from their surrounding tissue — known as the microenvironment — to tell them when to divide, when to rest, and when to die. Losing this dependence is one of the defining features of cancer. But in gastric cancer, researchers have long struggled to explain how this transition occurs.

This problem has been tackled by a joint international research team led by Dr. LEE Ji-Hyun, Dr. KOO Bon-Kyoung, and Dr. LEE Heetak at the Center for Genome Engineering within the Institute for Basic Science (IBS), in partnership with the laboratories of Prof. CHEONG Jae-Ho and Prof. KIM Hyunki (Yonsei University College of Medicine) and Prof. Daniel E. STANGE (TU Dresden / University Hospital Carl Gustav Carus). The team has identified a previously unknown mechanism that allows early gastric cancer cells to become self-sufficient. The findings provide a new framework for understanding how stomach cancer begins — and point to potential new targets for treatment.

Biochar can help turn agricultural waste into clean hydrogen fuel while cutting carbon pollution, according to a new study in the journal Biochar. Researchers report that a simple two stage catalytic system using corn straw, biochar, and nickel based catalysts more than doubled the hydrogen content of the gas produced during biomass pyrolysis and sharply reduced the carbon deposits that normally deactivate catalysts.

A public YouTube recording highlights how nature-based biomass solutions could deliver gigaton-scale carbon removal while supporting sustainable development

With advanced manufacturing tools, the team of materials science and engineering Associate Professor Hang Yu has created a shape-memory material that can be 3D-printed in bulk — a quest Yu began as a postdoctoral student at MIT.

Discover how space–ground fluid AI is set to revolutionize 6G networks by integrating edge AI with satellite systems, enabling global AI services and overcoming challenges of satellite mobility and communication rates. Explore the innovative techniques proposed in a new study published in Engineering that could shape the future of intelligent connectivity.