IRI-9527982 Wong, Stephen University of California-San Francisco $49,666 - 12 mos SGER: content-Based Retrieval of Multimodel Medical Images Medical images are at the heart of patient's diagnosis, therapy treatment, surgical planning, outcome assessment, and medical education. They, however, impose the most stringent requirements of information management in the future multimedia healthcare environments. Accordingly, this research will explore new techniques of content-based retrieval of multimodal medical images, including magnetic resonance imaging (MRI), positron emission tomography (PET), magnetic resonance spectroscopy (MRS), and magnetoencephalography (MEG), as well as complemented textual reports and image annotations. The term "content" is in a general sense, including structural and functional contexts of particular objects in an image, the objects themselves and their three dimensional anatomical structure, and even their semantic relationships across different kinds of medical images. The success of this exploratory research will provide key technical ingredients to construct future multimedia healthcare information systems and open up new vistas for digital medical library.