The long term objective of this research proposal is to develop a new class of medical information systems, i.e. image based medical information system, that can manage, process, and communicate vast amounts of multimodality medical images and their associated clinical textual data to support clinical practice, basic research, and education. The proposed research is designed to conduct a critical review of existing techniques in medical information management, develop an image data model, investigate algorithms for image processing, investigate software tools for organizing and indexing multimodality brain images by content, and develop a complete software package that incorporates the above tools. The brain imaging modalities involved are magnetic resonance imaging, magnetic resonance spectroscopy, positron emission tomography, and magnetoencephalography. The final stage will see the collaboration with clinical faculty in neurology and radiology to test the functionality and operational performance of the software package. The testing stage involves the application of the software package in managing multimodality medical images and clinical data for epilepsy surgical planning and multiple sclerosis diagnosis.
