Movement tracking of the carotid artery with 4D ultrasound imaging

Abstract: The leading cause of death in the world is cardiovascular diseases. Scientific discoveries have been made that correlate the movement of the arterial walls and blood flow patterns with cardiovascular diseases. These movements have been studied with ultrasound 2D imaging. From 2D movement tracking it is observed that the arterial wall moves in two directions, radially and longitudinally, and research has made connections from the patterns of movement to cardiovascular diseases. Blood flow patterns have been studied in 3D, with Doppler ultrasound and 3D MRI, and spherical blood flow patterns have been distinguished. This thesis is a first attempt to study the arterial wall movement of the common carotid artery with the use of 4D ultrasound imaging. To do this 4D ultrasound images were constructed by acquiring 2D motion images of the common carotid artery of three healthy volunteers, triggered with the use of ECG signals. These images were put through a 3D shift phase tracking algorithm developed for this thesis, using a program called MATLAB. To view the motion tracking of the arterial wall, a three-dimensional segmentation was created manually using 2D-slices. Applying the motion tracking to the coordinates of the segmented shape, a motion 3D image could be seen. It was not possible to draw any conclusions from this moving shape, so two other methods were used for validation of the 3D tracking algorithm. The first method used 2D slices, plotting a moving segmented slice on top of the corresponding ultrasound 2D image slice, to observe if the movement followed the original image. The second method compared the results in movement plots, plotting the 2D movement both radially and longitudinally and comparing this with an existing validated 2D shift phase tracking method. The results showed that the 3D motion tracking worked better in the radial direction than in the longitudinal direction, however neither results were completely correct. With many possible error sources, this method still seems promising, and one main issue is the resolution of the images, which needs to be increased for better results.