Methods for providing rural telemedicine with quality video transmission

Abstract

Telemedicine has been identified as a tool to distnllUte medical expertise to medically underserved rural community. However, due to the underdeveloped or non-existent telecommunication infrastructure, which is needed as the platform for telemedicine, the full benefits of telemedicine are yet to be realized in most parts of South Africa and Africa as a whole. This study aims to explore ways on how to provide lP-based lCI system that can be used as a communication platform for telemedicine in rural areas. In order to emulate the onsite face-to-face consultation experience, the rural telemedicine system must be able to provide quality video transmission. Quality video is also important in order for the physician at the distant end to be able to make correct diagnosis. Hence the main focus of this study is on ways ofproviding quality video over lP-based multiservice network. A conceptual model of a rural area network that can be used for rural telemedicine has been deVeloped, and different access technologies that can be used for rural areas are presented. Techniques for compesating IP best effort datagram delivery are provided. Factors that can affect the quality of video transmission on an lP-based packet network are identified, and a holistic approach to mitigate them is proposed. That includes adopting coding techniques that will provide coding efficiency, high quality video that is consistent at high and low bit rates, resilience to transmission errors, scalability, and network friendliness, which will result in perceived quality improvement, highcompression efficiency, and possibility of transportation over different networks. Secondly, it also includes mechanisms to compensate for packet networks idiosyncrasy, especially JP best-effort debilities, in order to meet the latency and jitter requirements of real-time video traffic. For video coding, H.264 is proposed as it meets most of the encoding requirements listed above, and for prioritising and protecting.video traffic from JP network's best-effort debilities a combination of differential services (DiflServ) and multi-protocol label switching (MPLS) have been adopted, where DiflServ is used for traffic classification and MPLS is used for traffic engineering and fast-rerouting in the event of route failure. To verify and validate the proposed solutions, modelling and simulation has been used, where the Network Simulator (NS-2.93) has been used to simulate network functions, and PSNR, VQM score and double stimulus impairment scale (DSIS) have been used for evaluating video quality.