Options
2006
Conference Paper
Titel
Case based reasoning based on content-based image retrieval approaches
Abstract
Based on demographic development and increasing life time in industrial countries, the time, a physician can deal with a patient will decrease dramatically over the next few years. Computer-Assisted Diagnosis (CAD) systems are one (technological) aspect for a possible solution to these problems in future health care system. Central element of our CAD prototype is a case-database which contains medical cases consisting of decisive images depicting objects and regions of interest as well as classifications for these objects. To access diagnostic knowledge in our database, we apply algorithms from Content-Based Image Retrieval (CBIR) and color texture analysis. Query point movement and dimension weighting methods are applied to steer the user's retrieval feed-back via a graphical user interface designed for non-specialist end users for easy access to the case database. The system was validated with four different comprehensive data sets. Additionally, simulations assuming perfect observes show that feedback iterations can improve the number of returned relevant cases significantly. The system behaves stable with increasing correctness up to a ratio of 20-30% wrong decisions per feedback iteration.