5G MEC+AI Pathology “Anti-Cancer Guardian”: Design of Intelligent Gastric Cancer Auxiliary Diagnosis and Warning Platform for Smart Hospital System

Purpose: Investigate the technical means of intelligent Gastric Cancer (GC) auxiliary diagnosis and warning, providing treatment guidance for physicians. Methods: The product is developed from a technical perspective, involving algorithm design, software development, and application design. Based on the UNet3+ digital pathology slices auxiliary diagnosis method, it enables precise segmentation of breast cancer pathology slices across the entire field of view, addressing issues such as excessive network complexity, high false positives, and inadequate capture of multi-scale information in existing algorithms. Additionally, utilizing a convolutional neural network-based cancer cell image classification algorithm enables rapid generation of predictive results for pathology slices. Key algorithms in 5G Mobile Edge Computing (MEC) for smart healthcare applications facilitate specific applications of 5G technology in integrated data exchange and device integration within medical consortia, achieving offloading of computational tasks and storage content to MEC nodes. This implementation includes patient physiological indicator warning mechanisms and device management techniques based on MEC nodes. The platform assists physicians in diagnosing more efficiently, aiding patients in earlier recovery, and providing robust support for treatment. Results: The new algorithm accurately segments and predicts entire cancer pathology slices, identifying cancerous regions. Conclusion: The “AI Pathology Anti-Cancer Guardian” intelligent GC auxiliary diagnosis and warning platform can analyze gastroscopy results more quickly and accurately, aiding physicians in diagnosis and treatment.