The
talk is focused on microstructure imaging paradigm which uses mathematical and
computational modeling together with cutting-edge machine learning and
parameter estimation techniques to facilitate development and application of
new biomedical imaging techniques. Our primary focus is on diffusion magnetic
resonance imaging (diffusion MRI) for medicine. Specifically, our techniques
enhance understanding and early detection of neurological conditions, such as
Alzheimer’s diseases, other dementias, multiple sclerosis, and psychiatric
disorders, and assist in the diagnosis and treatment assignment for solid
cancers. The paradigm uses model-based approaches to estimate tissue features
traditionally accessible only through invasive biopsy and histology. The
ultimate aim is replace such procedures, which currently underpin the
gold-standard diagnosis of most cancers and many neurological disorders, with
non-invasive imaging.