Some studies on hemodynamic modeling in human cerebral vasculature

Abstract

The arteriovenous malformation due to flow of blood through the carotid vascular system enlighten the need for study of hemodynamics in the same. Human cerebrovasculature consists of a set of connections of blood vessels which supplies blood to all the parts of the human brain. Hemodynamics plays an essential role in genesis and progression of some vascular diseases. The clear idea on patient specific carotid vasculature and hemodynamics there in, helps physicians to detect the disease like aneurysm etc. The main objective of this study is to segment the human carotid arteries from CT angiography images and do hemodynamic analysis to predict the anomaly in human cerebrovasculature. As the real medical data collections are so troublesome and patient permission is also needed, the real data are not only chosen, but also some phantom data are prepared to study the hemodynamic parameters thoroughly. Multiple tubular structures with varying diameters are combined using piece-wise cubic Bèzier curves to generate the complete arterial phantom, around the Circle-of-Willis which is presented at the base of the brain. The segmentation of vasculature from soft bones using FDT-based geodesic path propagation approach has been achieved and it successfully over performed some pervious work. Aneurysm is considered as a vascular disease as it is a very deadly disease to be taken care of. As the hemodynamics plays an important role in the progression and genesis of aneurysm, hemodynamics of cerebrovasculature is studied on different cerebrovascular structure with and without aneurysm. CFD based flow analysis is also applied to these cerebrovascular structures with and without aneurysm. Qualitative comparisons of CFD based hemodynamic parameters in human cerebrovascular structure with and without aneurysm are also performed. Digital flow has been generated to resemble the blood flow in vascular structure. Using the digital flow based model different hemodynamic parameters like velocity, wall shear stress have been generated and from those hemodynamic parameters and some flow based parameter the feature set has been designed which has been used to classify the anomalous region and normal region of the cerebrovascular structure.