Chair for Computer Aided Medical Procedures & Augmented Reality
Lehrstuhl für Informatikanwendungen in der Medizin & Augmented Reality

THIS WEBPAGE IS DEPRECATED - please visit our new website

Master Thesis: Registration of Multi-View Ultrasound with Magnetic Resonance Imaging

Supervision: Bernhard Fuerst, Ahmad Ahmadi and Wolfgang Wein

Freehand 3D ultrasound scans offer a high flexibility of the acquisition geometry and in particular offer the possibility to scan the anatomic target region from several viewpoints or perspectives. However, each 3D view is affected with its own linear and non-linear displacement errors and distortions, partly due to calibration and tracking inaccuracies and partly because of the variable speed-of-sound map in human tissue. One example is transcranial ultrasound scanning, in which the same deep-brain areas are scanned through both the left and right pre-auricular bone windows, with implications for improved early diagnosis of Parkinson's Disease. Together with our clinical partners at Klinikum Großhadern, we have collected a dataset of around 50 subjects, where such errors are present quite prominently. Alongside the 3DUS scans, however, we have also obtained MRI scans from all subjects. The objective in this thesis is to achieve a multi-channel US-US-MRI registration in 3D which will later allow to obtain an improved bi-lateral 3DUS reconstruction. For this, a novel registration method based on a current MICCAI publication from 2013 has to be extended in terms of theory and implementation. On a larger scale, the extension could have a profound impact on general multi-channel multi-modal registration problems.

Bi-lateral 3D transcranial US scans with optically co-registered MRI.
(Click on the image to see the fullsize version.)

For further information please contact:

Bernhard Fuerst:

Title: Registration of Multi-View Ultrasound with Magnetic Resonance Images
Director: Nassir Navab
Supervisor: Bernhard Fuerst, Wolfgang Wein, Ahmad Ahmadi
Type: Master Thesis
Area: Registration / Visualization, Medical Imaging
Status: finished
Thesis (optional):  

Edit | Attach | Refresh | Diffs | More | Revision r1.8 - 12 Aug 2013 - 09:11 - BernhardFuerst