n this thesis, we will work on medical image processing. More particularly, we will focus on Wall Motion Tracking in Intracranial Aneurysms by means of accurate and robust segmentation in a 4D Electrocardiography (ECG)-gated Computed Tomography (CT) image series. Detection of this barely- perceptible motion is a challenging task and has been given relatively little attention in the literature. Moreover, the existing state-of-the-art suffers from lack of accuracy and/or automation. Our goal is therefore to contribute to the improvement of the existing image segmentation and motion tracking models and methods by addressing this specific and difficult application. Since motion analysis is aimed at being used as the basis of vital decisions, its method must be validated using realistic phantoms. This is an important missing element in the recent studies, which we would also like to address. Intracranial aneurysm is a very dangerous and widespread problem. Fortunately, by the increasing use of medical imaging in different diagnoses, un-ruptured aneurysms are nowadays detected fortuitously while checking for other problems and may be treated in one of three ways: preventive surgery, neuro-radiological intervention or careful observation. At the moment, treatment depends mainly on the size and location of the aneurysm. According to studies [Wiebers et al., 2003], and considering the risks involved in surgery, careful observation is preferred for small aneurysms. However, a study in 2004 of 280 patients with ruptured aneurysms showed that a large proportion had small or very small aneurysms [Ohashi et al., 2004]. This means that at the moment there is no reliable means of predicting the rupture. Since observations during surgery shows that the aneurysms pulse, medical researchers are keen to study the relationship between the pulsatility and the rupture. However, there is no existing reliable method to detect this motion. We are therefore working on developing such a method for quantifying pulsation. Considering the fact that the motion is in the range of the noise or image artefacts, pre-processing is of utmost importance. Afterwards, we will focus on improving the state-of-the-art in accurate and robust segmentation. Finally, we want to go beyond the motion estimation methods, to improve current accuracy and automation. The following sections will present the state-of-the-art, the status of the present research, the detailed research plan and the schedule.