Purpose: Vessel-encoded (VE) pseudo-continuous arterial spin labeling (p-CASL) is a territorial ASL (T-ASL) technique to identify the perfusion territories of cerebral arteries. The aim of this study was to validate the output of three Vessel-encoded p-CASL image processing methods, k-means clustering with and without subsequent linear analysis and a Bayesian framework, by comparison with the perfusion maps acquired with super-selective p-CASL.
Methods: The comparison was done quantitatively using the Hausdorff distance and Dice similarity coefficient in the territories of the right and left internal carotid arteries, the basilar artery, and the right and left vertebral arteries. A qualitative comparison was done in the areas of the anterior and posterior circulation, and the deep gray matter.
Results: The overall agreement between the Vessel-encoded p-CASL image processing methods and super-selective p-CASL was good; with the difference that the linear analysis and the Bayesian framework were able to detect mixed perfusion.
Conclusion: Planning-free Vessel-encoded p-CASL with k-means clustering appears suitable as a general purpose T-ASL strategy, but to determine mixed perfusion a combination with linear analysis, or the Bayesian framework is preferable, which are superior in this regard. To accurately determine the perfusion territory of a single vessel, super-selective p-CASL is still recommended.
Keywords: anatomy; brain; function; neurological; normal.
Copyright © 2013 Wiley Periodicals, Inc.