Wireless Capsule Video Endoscopy For Population-Based Colon Cancer Screening Using 5G network

Abstract

We develop a battery-free communication system for a wireless video capsule endoscope with potential video streaming of a rate of up to 15 Mbps. We apply our innovative approach using backscatter for implants, a RADAR approach that remotely reads the information from the deep implants, such as the video capsule endoscope. We use the 5G deployed network with edge computing and slicing to transmit the video data from the capsule to a high-performance computing platform in a secure way with guaranteed end-to-end latency to perform polyp detection and localization. This way, the energy-intensive inference using deep learning neural networks for polyp detection and localization can be completed in the edge where control signals are sent back to the pill to increase the spatial and temporal resolution of the video, to obtain high-quality images for further analysis. The paper provides a system-level description, demonstrating the capsule data streaming to the network.