Amir Arsham Amini, Ph.D.

DEPARTMENT OF Internal Medicine
Keywords: imaging, MRI, quantitative analysis, image analysis, cardiac function, volumetrics

The efforts of our laboratory are focused on biomedical image processing, analysis and visualization, primarily the task of extracting information from medical images acquired from X-ray devices, MR imagers, nuclear medicine scanners and ultrasound. Analysis of this type of image has expanded rapidly in scope in recent years, driven in part by advances in parent disciplines of image formation and reconstruction as well as computer vision and shape modeling. However, expansion owes in part to the steadily larger availability of 3D image data which provides a playground for volumetric modeling of tissues, without dealing with projection operators encountered in optics. In order to extract information in medical images in a more automated, accurate and reproducible way, methods are devised that combine the underlying image physics with explicit or implicit data structures that effectively account for data originating from living tissues.

MRI tagging is a relatively new functional imaging method which has the ability to noninvasively and selectively alter tissue magnetization and create tagged patterns within a deforming body such as the heart muscle. The resulting patterns define a time-varying curvilinear coordinate system on the tissue, providing a mechanism for determining motion of material points at a sparse set of 3D points. Our laboratory is actively developing new mathematical and computerized image analysis methods for measurement of 3D non-rigid motion of myocardial tissue from this modality in addition to development of state-of-the-art computer graphics techniques for visualization of 3D anatomical and biological structures.

In collaboration with clinical colleagues, our laboratory is also interested in development of techniques for classification of tissues into normal and diseased from the tracked myocardial material points. Additional areas of research involve processing velocity data for flow analysis and visualization, segmentation and tracking of the left venticle and vascular motion from MRI, echocardiography and X-ray data sets. For more information, please visit our web site located at http://www-cv.wustl.edu.