The work presented specifically targets the accessibility to medical information. We used a French medical lexicon (specifically created for the medical domain), and built an index tool to particularly recognize a concept from a medical thesaurus that is present in a sentence written in natural language. First we indexed medical documents with a set of concepts and then demonstrated the utility of such indexing by developing a search engine and various tools which include: keyword identification, document similarity and automatic document synthesis. This indexing greatly aided in reducing the repository complexity of natural language documents. In addition, the evaluation results demonstrate that this indexing retains the main semantic information.
