This paper discusses results of the research in the area of artificial intelligence applications in telemedicine. The main goal of research is to manage multiple expertise obtained from experts-physicians in different countries to develop decision support medical system of broad earmarking based on telecommunication tools. The multilevel representation of medical data is discussed based on the apparatus of metastatistics. The technique is able to acquire semantically essential information from complex dynamics of quasi-periodical medical signals by applying recursively ordinary statistical tools. The voting-type technique is used to find consensus among medical experts in their description of medical knowledge. Research results are applied for developing Telediagnostic Expert Medical System and Medical Teleconsulting Support System.