Patient-specific vascular models for endovascular and open operative procedures Elizabeth Bullitt * , Stephen Aylward Departments of Surgery and Radiology, Division of Neurosurgery, University of North Carolina, 148 Burnett-Womack CB #7060, Chapel Hill, NC 27599, USA Abstract Object: Treatment of arteriovenous malformations (AVMs) and large base of the brain tumors is difficult partly because the surgeon must form a mental model of the lesion and its vascular supply. In planning treatment, it is important to localize supplying arteries, arteries of passage, and draining veins. This report provides an overview of software that offers the surgeon and interventional radiologist a detailed, three-dimensional (3D) map of each patient’s vasculature. A feature of our approach is that it defines vascular connectivity relationships so that one can estimate ‘‘safe’’ points of arterial occlusion. Methods: Vessels are extracted from 3D image data and symbolically linked to form vascular trees that represent blood flow patterns. These vessel representations can then be viewed in 3D in a program that allows the user to color-code connected groups of vessels or to selectively turn them ‘‘off’’ to either simulate catheter passage or to avoid obscuration. The 3D image data can also be aligned with digital subtraction angiograms (DSA) obtained from the same patient, in order to help the interpretation of the 2D image data by supplementing it with 3D information. Conclusions: Computer-assisted, three-dimensional visualizations of complex vascular anatomy can be helpful to the surgeon and interventional radiologist. Such programs can provide critical information not readily available by traditional techniques. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Magnetic resonance angiography; Digital subtraction angiography; Vessels; Intracerebral vasculature; Arteriovenous malformations; Brain tumors 1. Introduction Surgical planning for large base of the brain tumors and arteriovenous malformations (AVMs) is difficult partly because it is difficult to form a mental model of complex vascular anatomy in three dimensions (3D). 0531-5131/02 D 2002 Elsevier Science B.V. All rights reserved. PII:S0531-5131(02)01084-1 * Corresponding author. Tel.: +1-919-996-1374; fax: +1-919-966-6627. E-mail address: bullitt@med.unc.edu (E. Bullitt). International Congress Series 1247 (2002) 129 – 138