Prediction of Disturbed Flow by Factor Analysis of Carotid Bifurcation Geometry

Hemodynamics plays an important role in the development and progression of carotid artery atherosclerotic lesions. Certain aspects of vascular geometry, which mediates local hemodynamics, might be risk factors that increase a vessel’s atherosusceptibility [1]. To further evaluate this “geometric risk factor” hypothesis, the relationship between geometric features and hemodynamic quantities thought to typify “disturbed flow” was recently investigated [2]. Fourteen intercorrelated geometric features were initially extracted from MR images of 50 carotid bifurcations, and multivariate regression based on an a priori selection of a subset of four of these geometric features was used to identify two that were predictors of disturbed hemodynamics. Here, this work is extended to simultaneously analyze the combined role of all geometric variables using factor analysis.