This course will cover the fundamentals of image matching (registration) methods with applications to biomedical engineering. As the fundamental step in image data fusion, registration methods have found wide ranging applications in biomedical engineering, as well as other engineering areas, and have become a major topic in image processing research. Specific topics to be covered include manual and automatic landmark-based, intensity-based, rigid, and nonrigid registration methods. Applications to be covered include multi-modal image data fusion, artifact (motion and distortion) correction and estimation, atlas-based segmentation, and computational anatomy. Course work will include Matlab programming exercises, reading of scientific papers, and independent projects. Upon successful completion, the student will be able to develop his/ her own solution to an image processing problem that involves registration.
Prerequisites: 18-396 or permission of the instructor, working knowledge of Matlab, and some image processing experience.
This course is cross listed with 42-708 Special Topics: Registration in Bioimaging