A medical content based image retrieval system with eye tracking relevance feedback

Medical images are a key element in disease prevention, diagnosis, treatment and patient follow-up. The advent of 3D imaging equipment has increased the amount of medical images produced daily and software tools that are able to search and retrieve images are becoming popular, however improvements are still needed to close the gaps between the expert and the computer image representation. This article presents a software tool that integrates a Content Based Image Retrieval (CBIR) system with an implicit relevance feedback system that uses data gathered from an eye tracker. The gaze point can be used to infer regions of interest in the query image thus allowing for searches based on global or local features, and to steer the retrieval process of relevant images. A preliminary evaluation of the system is presented and discussed.