Contextual medical image viewer.

Author(s)

Ramon Alfredo Moreno

• Document type: Doctoral Thesis
• Press: São Paulo.
• Institution: Universidade de São Paulo (USP). Escola Politécnica (EP/BC)
• Defense date: 2005-04-15

Examining board members

Sergio Shiguemi Furuie; Marco Antonio Gutierrez; Paulo Mazzoncini de Azevedo Marques; Agma Juci Machado Traina; Marcelo Knorich Zuffo

• Advisor: Sergio Shiguemi Furuie

Abstract

The Electronic Health Record should, theoretically, aggregate all possible sources of information from a patient, no matter their location or storage format, and allow authorized people to access this information. However, in practice, the medical information systems are heterogeneous and hard to integrate, besides the fact that medical information is complex and poorly structured. To handle this problem the Heart Institute of São Paulo leaded a project named PACS-HC with the aim of integrating medical information from heterogeneous systems, focused mainly on medical images. This work proposes the visualization component of the PACS-HC, called Contextual Medical Image Viewer. The proposed viewer is capable of performing automatic actions using contextual information, like user role, location and image dimension. Some examples of actions are the retrieval of similar images from a case image database and reference information from the WEB. The contextual viewer uses a context model, implemented as an ontology, to perform those actions. It also relies on an extensible architecture for supporting the processing and visualization of medical images. The result is the software ODIN, which is capable of displaying images from different modalities and dimensions and using contextual information to perform actions using rules. It can be concluded that the use of contextual information is possible and desirable as a way to improve user interaction and it must be added that there is an extra potential for the use of medical information, not only image-related, but to improve the healthcare attention. As future work there are investigation of new ways of using contextual information, such as through intelligent agents, and the refinement of the context model from use cases that reveal the needs of physicians and healthcare professionals in different environments and situations. (AU)