Ramon Alfredo Moreno BEng, Abstract Co-Author: Nothing to Disclose

Sérgio Shiguemi Furuie PhD, Presenter: Nothing to Disclose

The purpose of this work is to develop a viewer of medical images that uses contextual information to perform actions that can help the user’s interaction with the software, for example, bringing a rank of similar images to the one being observed if the user is a medicine student.

The viewer is being developed using Java and has an architecture based in the Mediator pattern to perform contextual actions.The viewer has an Interface Manager, which receives Subscribers for different kinds of events, e.g., the opening of a new image. When the event occurs the Interface Manager contacts all the subscribers asking for the actions they intend to do. After analyzing the actions the Interface Manager decides to perform one or more actions. Then it notifies the Subscribers of the status of the actions: performed, canceled or erroneous.The Interface Manager is an abstract Singleton that instantiates one of its subclasses. In this way it is possible to change the behavior of the viewer.The Subscribers define their actions by means of an XML string. The XML is composed of tags that are processed to perform an action. For example, a tag “displayViewer” will show in the main frame of the viewer an instance of the class specified in the parameter “className”.The tags of the XML string can be customized to perform new actions, extending the class Tag.

The viewer is being developed at the Heart Institute of São Paulo. A prototype version is already available for use in the hospital and can perform the viewing of 2D, dynamic 2D and views of 3D sets. The viewer also comprises some tools for processing and manipulation of the images. The framework developed allows the extensibility of the viewer making easy to introduce new tools for processing and define new context actions.

The proposed model is a step towards the creation of an environment where the context can be used to improve the user capabilities.The architecture is extensible allowing the inclusion of new tools and viewers and able to grow with the contribution of other developers and reveal new uses for the contextual information in different areas of visualization and processing of medical images.

Moreno, R, Furuie, S, Contextual Medical Image Viewer.  Radiological Society of North America 2004 Scientific Assembly and Annual Meeting, November 28 - December 3, 2004 ,Chicago IL. http://archive.rsna.org/2004/4411871.html