Blood Flow Simulator using Medical Images without Mesh Generation

Abstract

Conventional blood flow simulators needed to generate meshes, which had to have fine resolution at the segments where significant flow pattern changes occurred. So, to generate ideal meshes, excellent ability in mesh generation itself was essential, at the same time, detailed knowledge in flow analysis was important. These had been reasons why the flow simulation was not widely used in medical field. We developed a new blood flow simulator without mesh generation. Our system used voxel model as simulation. The voxel models were directly made from grey scale medical modalities such as CT-Angiography, MR-Imaging or 3D-Angiography. We applied our system to two models, one was the model of coil embolization for brain aneurysm, the other was protection for carotid artery stenting. In coil embolization model, flow pattern change in aneurysm sac were calculated giving various amount of coils. Increasing volume embolization ratio, flow velocity in the sac was decreased. If protrusion of coil crossing flow in the parent artery, flow in the sac possibly increase. In case of carotid artery stenting, protective balloon occlusion of distal internal carotid artery could not prevent migration of debris from the site of balloon angioplasty to external carotid artery. In coil embolization, our simulator successfully showed decrease of flow velocity in aneurysm sac depending on increasing amount of coil(s). The result was significant alert that the crossing protrusion of coils possibly increase flow in the aneurysm sac. At carotid artery stenting, debris to the ECA might cause brain ischemia because of possible ECA ICA anastomosis. Our simulator can work stably even extremely complex structure such as “coils in aneurysm” model without detailed knowledge on the computational fluid dynamics nor mesh generation. Moreover, voxel models were directly built from medical modalities without special knowledge. Here, our simulator can work well in practical medical field.