Researchers from Donghua University and Yanshan University have developed an enhanced neural network called EF-YOLO, which significantly improves the automated detection of colorectal polyps—a key precursor to colorectal cancer. By integrating advanced modules such as transformer, deformable convolution, coordinate attention, and a newly designed multi-scale aggregation mechanism, the model achieves a mean average precision (mAP) of 96.60% on standard medical datasets. This represents a major step forward in reducing missed detections during colonoscopy and supporting early clinical intervention.

Modern large-scale warfare imposes extreme psychological stress on soldiers at the front. Ukrainian soldiers are being trained in Norway to help better protect them from this trauma.

Inspired by the multimodal perception characteristics of insect antennae (for force, temperature and gas), the ECUST interdisciplinary team developed the Electronic Multifunctional Sensing Antenna (EMSA)—a flexible thin-film self-decoupling multimodal sensor fabricated via a maskless laser direct writing process. The sensor realizes intrinsic decoupling of temperature and strain, as well as simultaneous thermo-mechanical-gas trimodal monitoring, with four core technological breakthroughs.

A recent study by University of Notre Dame researchers describes a promising alternative for next-generation cooling using thermoelectric technology, which has no moving parts ad no gaseous refrigerants, allowing for zero leaks.

A team of gas sensors led by Ronghan Wei from Zhengzhou University in Zhengzhou, China, synthesized In₂O₃ with oxygen vacancies (In₂O₃-L) and anchored Pt single atoms, enabling real-time detection of low-concentration acetone. The In₂O₃-L-Pt sensor demonstrated a 5-fold improvement in response to 20 ppm acetone at 220 °C compared to In₂O₃, exhibiting rapid response/recovery times (2/30 s), ultra-low theoretical detection limits, excellent selectivity, and outstanding long-term stability.

Artificial intelligence systems are becoming increasingly powerful—but also harder to understand.