A patient-specific computer model for prediction of clinical outcomes in the cerebral circulation using MR flow measurements

Abstract

A patient-specific, computer model of the cerebral circulation to predict cerebral blood flow and pressure is presented. The model is based on a previously reported numeric model consisting of a network of distensible vessels with pulsatile flow. The enhanced model uses a sector scheme to determine the efferent resistance distribution for a specific patient. An iterative algorithm was developed to determine the patient-specific efferent resistance distribution from in vivo cerebral blood flow measurements obtained using phase contrast magnetic resonance angiography (PCMRA). In comparison with PCMRA flow measurements and clinical outcomes, the enhanced model shows its ability to predict cerebral flow well in three patients who underwent a balloon occlusion of the carotid artery. A model that accurately predicts cerebral blood flow for different treatment scenarios can provide the surgeon with an invaluable tool in the management of complex cerebral vascular disorders.