Arjen Lindenholz

11 General Introduction 1 Challenges in visualizing the intracranial vessel wall Compared with extracranial vessel wall sequences, several unique challenges can be identified for intracranial vessel wall MRI. Chapter 3 provides an overview of these challenges in technical requirements and assessment of these images. First, most early MRI sequences were ‘copy-pasted’ from extracranial vessel wall MRI as two-dimensional (2D) acquisitions with anisotropic voxel size. 14,16,17 Although this does not pose significant issues in the larger extracranial arteries due to their mostly straight configuration, it caused hindering partial volume artefacts in the smaller, more tortuous intracranial arteries. Another drawback of 2D acquired vessel wall MR images is that the intracranial vessel wall and atherosclerotic plaque cannot be assessed in multiple planes without necessitating acquisitions in all three directions (axial, sagittal and coronal), tripling the total acquisition time. 3D intracranial vessel wall MRI sequences are therefore gaining popularity, allowing for isotropic acquisition which renders multiplanar reformatting of the vessel wall images possible. However, (in-plane) spatial resolution of these sequences is generally lower than of the 2D sequences. 15 Second, a high contrast-to-noise ratio (CNR) is required to delineate the thin intracranial vessel wall from the surrounding intraluminal blood and extraluminal CSF. Black blood techniques are used in extracranial and intracranial vessel wall MRI to suppress the signal from the intraluminal blood. Whereas the extracranial arteries are surrounded by fatty and other tissues, the intracranial arteries are also surrounded by variably flowing cerebrospinal fluid (CSF). This means that, in addition to black blood techniques, the signal from the surrounding CSF also needs to be suppressed. Initial intracranial vessel wall MRI sequences had difficulty discerning the vessel wall from both the arterial lumen (because of the different flow directions of the tortuous intracranial arteries) and from the surrounding CSF. 5 A variety of (flow saturation) techniques have been developed to suppress the MR signal from blood and CSF, mostly relying on the flow properties of luminal blood (black blood sequences) and the use of specific preparation pulses (double inversion recovery pulses), albeit with varying success in improving the delineation of the intracranial vessel wall. 2,15,16,18,19 Third, although the technical developments have made clear visualization of the small and tortuous intracranial vessel wall possible with improved spatial resolution, SNR and CNR, an important drawback of all these highly-demanding requirements are the relative long acquisition times. Long scan times induce motion artifacts, for example when examining stroke and TIA patients who are often elderly with moderate to severe neurological disability and therefore prone to movement. This also limits the extent of atherosclerotic plaque characterization that is possible. For instance, a choice often has to be made between acquiring a T 1 -weighted vessel wall sequence before and after contrast administration,