Global Intelligence Database
New research challenges the idea that vision is a one-way street, showing instead that language-based knowledge actively shapes how we perceive the world. By comparing AI models with brain scans, scientists found that abstract meaning acts as a scaffold to organise raw visual inputs.
By Zhang, Tu, Yin, Zhang, Shi, Li, Dai, Hao, Jiang
Artificial intelligence shows significant potential to make cosmetic surgery more objective and predictable, from simulating outcomes in 3D to monitoring recovery. However, a major review reveals the technology is still in its infancy, with most studies being early-phase and requiring far more rigorous validation before widespread clinical use.
By Shoham, Naveh, Confino, Zoabi, Govrin, Fliss, Barnea
Researchers have developed ResTab Net, a new AI system that combines tissue images and protein data to diagnose breast cancer with 98.56% accuracy. Unlike traditional 'black box' models, this system uses explainable AI techniques to show clinicians exactly how it reaches its diagnostic conclusions.
By Jaikumar, Praveena
A new meta-analysis examines the shift from rule-based to generative AI in mental healthcare. The findings suggest moderate effectiveness in reducing depression and anxiety, particularly when the AI focuses on social interaction rather than specific tasks.
By Zhang, Zhang, Xiong, Sui, Tong, Lin
A comprehensive new review explores how machine learning and sensor fusion can overcome data scarcity to reliably detect falls in older adults. By bridging engineering and gerontology, the study outlines pathways to improve safety for an ageing population.
By Gattani, Dixit, Patil, Gupta, Navghane, Hule, Srinivasan