Architecture and applications of the parallel distributed image search engine (ParaDISE)

Abstract

Medical images are important to physicians for diagnosis and treatment planning. The number of images produced in medical institutions bas been increasing rapidly over the past years. Making these imaging data available and allowing medical professionals to perform retrieval based on visual characteristics of images is the challenge that content-based image retrieval systems are tackling. In this paper we propose a system called ParaDISE (Parallel Distributed Image Search Engine). It allows indexing and retrieving images using both visual features as well as text. It is designed to scale well with big data sets, and it should be easily extensible with new features. The architecture of the system is described and its use in medical applications is presented. An open source release of the system is in progress.