Proposition of a new classification of the cerebral veins based on their termination.

The existing classifications of cerebral veins have certain problems, including limited adequacy to uniquely describe neurovascular networks in three dimensions (3D), mixture of deep and superficial veins, and ambiguity of territories-based parcellations as veins may course on multiple territories. Classification discrepancies exist in subdivision, region drained, and parcellation criteria. Recent developments in diagnostic imaging and computers enable to acquire, create, and manipulate complete vascular networks, which also call for a new classification of cerebral veins. We propose a new classification suitable for the description of the complete cerebral veins, providing a clear separation of the superficial cortical veins from deep veins, and facilitating presentation and exploration of cerebral veins in 3D with respect to surrounding neuroanatomy. It is based on terminating vascular subsystems (rather than draining regions). It divides the cerebral veins into cortical, deep, and posterior fossa veins. The cortical veins are subdivided into two groups: terminating in dural sinuses and terminating in deep veins. The posterior fossa veins are subdivided also into two groups: terminating in dural sinuses and terminating in deep veins. This classification was illustrated with a cerebrovascular model containing over 1,300 vessels. This new classification has many advantages. It is simple, clear and didactically useful; avoids mixture of superficial and deep veins; shows overall hierarchical structure and topographical relationships including tributaries; is useful in analysis of 3D vascular trees extracted from imaging; and may be used in conjunction with the existing parcellations.