Registration of the lungs in thoracic CT images is required in many fields of application in medical imaging, for example for motion estimation, analysis of pathology progression or the generation of shape atlases. In this paper, we present a robust registration approach that has been optimized for the registration of thoracic CT data. The algorithm con-sists of an initial shape-based adjustment of lung surfaces followed by an intensity-based diffeomorphic image registration. The approach is evaluated based on 20 CT scans provided for the EM-PIRE10 study for pulmonary image registration. A fourth place out of 34 participants suggests a good applicability for the registration of lung CT images.
|Publication status||Published - 2010|
- Beijing, China
Duration: 20.09.2010 → 24.09.2010
|Period||20.09.10 → 24.09.10|