Development of a computer program for volume estimation of arteriovenous malformations from biplane DSA images

Abstract

The goal of this work was to develop a computer program for accurate estimation of size and location of intracranial arteriovenous malformations (AVM) from biplane digital subtraction angiography (DSA) images. The program will be used for diagnostic purposes to predict the outcome of Gamma knife radiosurgery and thus give basic data for the optimization of the management and comparison of different treatment modalities. Our solution is based on the so called intersecting cone model (ICM). Volume measurements with this model have been shown to correlate well to volume data from the dose-planning equipment (Leksell GammaPlanr®, Elekta, Sweden). The method described in this paper produces a fast and accurate implementation of the ICM suitable for clinical practice. The implementation has been validated from phantom images and simulated images and the errors have been shown to be small compared to the approximations made by the ICM.