Content-based Searching of Medical Images Using a P2P Network

Abstract

Acontent-based searching method of medical images using a P2P network was studied. Medical images are stored in peers together with the diagnostic findings (symptoms, etc.). The keywords of the diagnostic findings are registered in index servers. The keyword candidates of the diagnostic findings relating to the given searching keywords (medical subjects, parts of the body or kinds of medical images) are found by a search agent and are displayed. The search agent learns so that the keywords are shown in the higher order of relevance to the given keywords. Medical images, therefore, can be searched more efficiently.