Tomographic techniques provide powerful tools for non-invasive imaging with various applications from clinical diagnosis to non-destructive testing. Since they acquire indirect measurements, the searched-for image has to be reconstructed from the collected data by solving an inverse problem.
This reconstruction process is well understood if the object under investigation ist stationary during the data acquisition. Imaging structural and functional changes of a specimen represents a challenging problem. Most modalities record the data sequentially, i.e. temporal changes of the object lead to inconsistent measurements. Consequently, suitable models and algorithms have to be developed in order to provide artefact free images. Following a detailed introduction to this type of imaging problems, we discuss two reconstruction approaches incorporating different types of motion information. We first exploit an explicit motion model and derive a suitable direct reconstruction method. Further, we present an iterative strategy which treats the dynamic behavior as uncertainty in the forward model. Both strategies are validated for data sets from computerized tomography with different dynamic behavior. Collaborators: Stephanie Blanke (Universit¨at Hamburg), Melina Kienle Garrido (University of
Stuttgart) and Anne Wald (Saarland University)
The talk will be given via ZOOM. Please send a message to firstname.lastname@example.org, so that we can send you an inivation for the seminar.
Angelegt am Wednesday, 07.10.2020 17:47 von Claudia Giesbert
Geändert am Tuesday, 17.11.2020 10:23 von Frank Wübbeling
[Edit | Vorlage]