In this paper, we design and implement a real-time monitoring system based on CPS (Cyber Physical System) for telemedicine in disaster areas where the infrastructure is destroyed. In this situation, our monitoring system relies on a wireless sensor network (WSN) in ad-hoc mode, which provides a very low data rate. There are several types of real-time traffic to transmit over the low-rate WSN: voice, bio data, and pictures of victims or disaster scenes. First, network quality is judged by the delay of voice packets. Also, priority is given to voice packets over bio data or pictures, because they are critical for telemedicine. Voice packet sampling is performed at intervals of 100ms, because voice packets must arrive at the receiving side in 100ms at least. According to the network quality, one image is segmented in a different size. As a result, real-time voice, bio data & segmented pictures can be transmitted in a low-rate WSN network in 100ms. Furthermore, the health diagnosis center is able to identify the site conditions. Through this study, if the system is applied to a real disaster situation, it will be effectively utilized for rescue activities by rescue workers.