Development of segmentation techniques for high-speed ultrasound imaging of cardiac arrhythmias

• Κύριος συγγραφέας: Κτιστάκης, Αθανάσιος
• Άλλοι συγγραφείς: Ψαράκης, Εμμανουήλ
• Μορφή: Thesis
• Γλώσσα: English
• Έκδοση: 2018
• Θέματα: Segmentation; Cardiac arrhythmias; Ultrasound; Κατάτμηση; Υπέρηχοι; Καρδιακές αρρυθμίες; 616.128 075 47

Cardiac arrhythmias are among the most common heart diseases affecting every year a large percentage of the population. 4D ultrasound imaging has become a popular technique both for the study of cardiac wall motion and in understanding deeply the mechanisms involved in cardiac arrhythmias. In this study, experimental data was obtained from a Langendorff-perfused pig heart and was subjected to preprocessing, segmentation and registration techniques. As a preprocessing step, filtering using curvature anisotropic diffusion was applied to decrease the speckle noise while enhancing the boundaries of the heart and preserving the edges of the 3D image. Subsequently, image segmentation using the Snakes algorithm along the different slices was performed for the reconstruction of the three dimensional model. To measure the local deformations on the cardiac wall image registration was applied. B-spline registration and Demons registration were examined both on the unprocessed data and on the preprocessed segmented models