| Abstract [eng] |
Automated segmentation of the abdominal aorta - an important and relevant problem that, once solved, can significantly optimize the work of medical professionals. Currently, doctors spend numerous hours manually marking the abdominal aorta in computed tomography images. The goal of this study is to investigate the characteristics of the abdominal aorta in computed tomography images, propose an automated method for its identification and segmentation, including aneurysms, based on the obtained insights, and review the encountered problems during the work along with their solutions. Throughout the study, relevant literature related to the automated segmentation of the abdominal aorta is examined, a diverse set of patient images is analyzed to identify features that will be used to develop the methods. A fully automated method for the identification and segmentation of the abdominal aorta and aneurysms is formulated and tested, and the obtained results are described with accompanying examples. |
