Enhancing Electronic Patient Record Functionality Through Information Extraction from Images

Complex multimedia medical information is being increasingly used to enable diagnosis. Despite the recognized potential of emerging exploitation of such medical information, known electronic patient record applications do not offer the possibility to make use of them. In order to examine the interest of handling complementary medical information, this work focuses on the integration into the electronic patient record of segmented lesions (e.g. tumors) and prior anatomical knowledge, in a neurological examination scenario. Based on a generic architecture, the proposed application proposes a rich and flexible graphic user interface, capable of simultaneously displaying 2D and 3D images along with segmented lesions, associated to anatomo-functional knowledge. Open source technology was chosen for prototype implementation. The main modules are described and analyzed