MRI-Based Modeling of Vascular Territories and Collateral Circulation of the Brain

J. R. Cebral, F. Calamante School of Computational Sciences, George Mason University, Fairfax, Virginia, United States, Institute of Child Health, University College London, London, UK, United Kingdom Introduction: A key aspect for the management of stroke patients is the differential diagnosis of hemodynamic border zone infarction and thromboembolic ischemia, which is largely based on the location of the visualized infarct relative to the arterial territories and border zone regions [1]. However, it has been shown (using unfixed human brain specimens) that the variability of the major cerebral arterial territories can vary quite significantly [2]. Since individual arterial territories are difficult to identify in vivo, interpretation of acute stroke subtype and underlying pathogenesis, based on the topographic patterns alone, is often inaccurate [1]. The leptomeningeal anastomoses (LMA), a connecting branch between two major arterial territories, is believed to be important in the understanding of stroke mechanisms. However, there is no consensus in the literature regarding the role of LMA in the outcome of the penumbra [3]. Therefore, there is a need for the development of new techniques for the in vivo mapping of arterial territories, as well as for the assessment of the changes to the border zone regions under compromised situations, such as carotid stenosis or occlusion. Furthermore, the development of models to investigate the effect of LMA on the arterial territories and the border zone regions could prove very beneficial; they could be used to further our understanding of the potential effect of LMA on cerebral circulation, as well as shed some light on their functional significance in stroke mechanisms. The currently limited understanding of the collateral circulation in the human brain can be greatly enhanced through the development of vascular models, providing the basis for future therapeutic and prognostic applications. This study presents a novel method to create a subject-specific vascular model, which is based on combining MRI and MRA data with computer generated 3D arterial tree structures.