seminar: Discrete tomography and its applications in biomedical imaging
speaker: Kees Joost Batenburg (University of Antwerp, Belgium)abstract: Tomography is a powerful technique for both in-vivo and ex-vivo biomedical imaging. Although modern CT-scanners are much faster than their predecessors, harmful effects of using X-rays are still a major concern. In particular, this is still a problem for laboratory experiments in medicine research, where scans on small animals have to be performed on a regular basis. The field of discrete tomography focuses on the reconstruction of objects that consist of only a few different materials (or, in the case of medical imaging, "tissues"). If each of the materials has its own characteristic grey value in the reconstruction, prior knowledge of these grey values can be used within the reconstruction algorithm. By using this prior knowledge, high quality reconstructions can be computed from significantly fewer projections than for conventional CT. A major additional advantage is that the reconstructed image does not have to be segmented, as it already contains only a single grey level for each material. Until recently, no discrete tomography algorithms have been available that can be used effectively on large experimental CT datasets. We will present a new iterative discrete reconstruction algorithm that is very fast and is capable of dealing with real experimental data. Our experimental results show that for suitable medical imaging applications, significant dose reductions can be obtained while maintaining reconstruction quality.
timetable:
Thu 18 Oct, 14:00 - 15:00, Aula Dini
<< Go back