Computational sciences and nuclear medicine imaging

点击翻译按钮获取中文摘要

Nuclear medicine imaging techniques, with their molecular characteristics and exceptional sensitivity, play an important role in modern medicine. By providing in-vivo functional information about the biodistribution of tracer molecules labeled with radioactive isotopes, they allow not only for identification of altered anatomy and function to help diagnose disease, but also for locating molecular receptor sites, studying biological markers expressed by diseased cells, and monitoring the presence and extent of specific disease processes. The main objective of the research program of the Medical Imaging Research Group (MIRG) is to use advanced physics, mathematics and computer science (jointly called here "Computational Sciences") to develop new and improved methods for diagnostic nuclear medicine, ones suitable for use both in everyday clinical practice and in medical and molecular research. MIRG's research is in several closely related areas: quantitative imaging, dynamic functional imaging, data analysis and visualization, and the investigation of novel imaging techniques. Some of our future goals include: (1) the development of an exact dosimetry technique based on our unified quantitative SPECT method (qSPECT) combined with an automatic 3-D volume segmentation routine (level-set based and similar methods) and detailed Monte-Carlo calculations of internal dose due to energy deposition by gamma photons, X-rays and electrons, (2) further research on the application of modern mathematical optimization techniques with different regularization schemes to reconstruct a multidimensional series of dynamic images, (3) development of software, construction and testing of a Compton camera. The new generation of cameras, based on detection of Compton scattered photons and using solid state detectors can potentially provide means for truly molecular imaging with sub-millimeter resolution (with the goal to image particular cells).