The goal of this research is to develop an open and extensible software environment for medical image segmentation. This environment will contribute to the public and scientific interest in at least three ways: (1) improved and efficient segmentation of medical images for various applications, (2) efficient creation of new image segmentation algorithms, and (3) improved evaluation for medical image segmentation algorithms. Image segmentation has many applications in medical imaging; however, it's use in current clinical practice falls far short of its potential. Commercially available tools are either designed for very specific applications, or are general-purpose image processing packages with little support for image segmentation. The proposed environment will be devoted to medical image segmentation and will have an extensible and open architecture. The extensible architecture will allow easy customization for specific applications and will also allow users to add their own algorithms. The open architecture will make the software platform-independent and will allow easy integration with existing applications, such as databases, analysis packages, or visualization tools. This environment will also provide tools for evaluation of medical image segmentation algorithms. To design such a software environment, we will use the latest innovations in design such a software environment, we will use latest innovations in software technology, such as object-oriented design and distributed objects.
We envision two types of users for this software environment--(1) medical imaging researchers or medical imaging solution providers, and (2) clinicians or clinical researchers. The first type of uses will be able to customize the application completely. The second type of users will use the software for specific applications. Our building-block approach to the design of the environment will allow rapid customization for different applications and for the different types of users.
