Abstract : Transmissions of multimedia data in emergency context such as, the occurrence of trauma or stroke are sensitive to the network delay, throughput and packet loss. Depending on the urgency, the communication channel quality can decide the survival of a patient. For saving lives, network infrastructure offering good Quality of Service (QoS), high broadband and low communication cost must be considered. In this context, Cognitive Radio Networks (CRN) seem appropriate. Indeed, this technology is well suited with its sensing ability and the scanning of a wide range of technologies, for assigning adequate frequency bands which meet the application constraints. In this paper, we propose a CRN-framework and an associated analytical tool that sets and imposes constraints to be respected, for ensuring reliable and high quality end-to-end communications.