The team from Northeastern University in the United States has successfully developed a new type of network-based artificial intelligence (AI) system. The accuracy of the system in detecting breast cancer is as high as 99.72%. The relevant research paper has been published in the latest issue of the journal Cancer. The team created an integrated deep learning model that can integrate and utilize various other models to improve diagnostic accuracy and reduce errors. Then, they trained these models using publicly available databases, including images of malignant and benign breast tissue stored in the Cancer Tissue Pathology Database (BCHD). The team stated that their newly developed AI system will view high-resolution images and learn how to identify cancer and diagnose it from historical data. The system almost never misses a tumor during biopsy and can maintain stable performance even after continuous diagnosis of multiple people, without affecting accuracy due to fatigue. Ideally, this system not only helps doctors identify patients faster and more accurately, but also helps to generate new AI models for diagnosing rare cancers that lack patient data. At the beginning of this year, the team launched an AI tool that could diagnose prostate cancer faster and more accurately. They hope to create a larger system based on these systems, allowing doctors to use these innovative AI technologies to diagnose a range of cancers. Researchers say that the new tool will "redefine digital pathology.". (Lai Xin She)