An architecture for implementing customizable medical image processing systems.

Abstract

Monolithic image processing systems containing a superset of imaging algorithms are difficult to use and require specialized knowledge of image processing. Thus they increase the workload of medical personnel instead of making the work situation easier. Customizable medical image processing systems on the other hand may be easily adapted to address various problems in the medical image processing domain integrating only the necessary subset of image processing functionality presented in on intuitive way. In this work we present an architecture for creating customizable image processing systems for the medical domain. We address three major topics: 1.) easy, goal-oriented customization of imaging systems by using a generalized algorithm model and repository, 2.) dynamic, data-oriented parameterization of the selected algorithms and 3.) semi-automated generation of user interface components for each new algorithm to be inserted in an imaging system based on cognitive ergonomics. We conclude with the presentation of an initial implementation of the architecture in form of an object-oriented framework for the creation of components for customizable medical imaging systems.