BEGIN:VCALENDAR BEGIN:VEVENT SUMMARY:Novel Theoretical Characterization and Optimization of Experimental Efficiency for Diffusion MRI (Ph.D. Defense) DESCRIPTION:Speaker: Divya Varadarajan, Electrical Engineering, University of Southern California Talk Title: Novel Theoretical Characterization and Optimization of Experimental Efficiency for Diffusion MRI (Ph.D. Defense) Series: Medical Imaging Seminar Series Abstract: Diffusion MRI (dMRI) has the unique ability to noninvasively quantify the Brownian motion characteristics of water molecules, and thus infer structural tissue features at microscopic spatial scales that are otherwise inaccessible through conventional millimeter-scale MRI. When applied to the brain, dMRI has emerged as an especially important tool for quantifying tissue microstructural characteristics that change as a result of factors such as brain development, plasticity, and pathology. Diffusion MRI is also used for reconstructing the white matter pathways that connect different brain regions.\n \n However, long scan times continue to be a major challenge for dMRI . While spending an hour or more in the scanner is acceptable forex-vivo tissue analysis or for motivated, healthy volunteers, it becomes challenging for in-vivo analysis for a number of important subject populations (such as younger children and sick individuals). In practice, scan times are constrained for human subjects, and studies have to make inferences working with the few samples acquired. The overall performance is directly impacted by the efficiency of the dMRI protocol, which consists of the sampling scheme used to acquire dMRI data, as well as the parameter estimation method used to make inferences from \n measured data.\n \n This work addresses the problem of making dMRI experiments as efficient as possible, including: (i) optimizing the scanner sampling protocol to maximize the amount of relevant information contained in the data, (ii) optimizing the parameter estimation protocol to maximize the amount of information extracted from the measured data, and (iii) developing statistical models that can be used to suppress noise contamination.\n Host: Prof. Justin Haldar DTSTART:20171004T100000 LOCATION:EEB 132 URL;VALUE=URI: DTEND:20171004T110000 END:VEVENT END:VCALENDAR