Symbolic description of intracerebral vessels segmented from magnetic resonance angiograms and evaluation by comparison with X-ray angiograms

We describe and evaluate methods that create detailed vessel trees by linki ng vessels that have been segmented from magnetic resonance angiograms (MRA ). The tree-definition process can automatically exclude erroneous vessel s egmentations. The parent-child connectivity information provided by our ves sel trees is important to both surgical planning and to guidance of endovas cular procedures. We evaluated the branch connection accuracy of our 3D ves sel trees by asking two neuroradiologists to evaluate 140 parent-child conn ections comprising seven vascular trees against 17 digital subtraction angi ography (DSA) views. Each reviewer rated each connection as (1) Correct, (2 ) Incorrect, (3) Partially correct (a minor error without clinical signific ance), or (4) Indeterminate. Analysis was summarized for each evaluator by calculating 95% confidence intervals for both the proportion completely cor rect and the proportion clinically acceptable (completely or partially corr ect). In order to protect the overall Type I error rate, alpha-splitting wa s done using a top down strategy. We additionally evaluated segmentation co mpleteness by examining each slice in 11 MRA datasets in order to determine unlabeled vessels identifiable in cross-section following segmentation. Re sults indicate that only one vascular parent-child connection was judged in correct by both reviewers. MRA segmentations appeared complete within MRA r esolution limits. We conclude that our methods permit creation of detailed vascular trees from segmented 3D image data. We review the literature and c ompare other approaches to our own. We provide examples of clinically usefu l visualizations enabled by our methodology and taken from a visualization program now in clinical use. (C) 2001 Elsevier Science B.V. All rights rese rved.