Hojjat Emami

Reducing the workload of physicians in telemedicine increases the accuracy of diagnosis and leads to effective treatment. This paper presents the theoretical foundations of an intelligent system for clustering patient reports in telemedicine. The proposed approach is based on natural language processing and text clustering methods. The proposed approach takes as input the patient reports written in unstructured natural text and structured tables, and groups them into some disjoint clusters according to their similarity in semantics. The proposed approach consists of four main steps: pre-processing, semantic analysis, similarity computation, and grouping. The incentive mechanism of the proposed approach is to semantically analyze and enrich the content of reports using distant ontology to improve the clustering performance. The experimental results on a benchmark dataset drawn from Ohsumed collection show that the proposed approach outperformed its counterpart method.