← Back to Book Detail

College of Engineering (16/39) -- RANGE: Undergraduate Research Journal

Browse
41%

College of Engineering

College of Engineering 17 Investigating Curvature Induced Circumferential Drifts to Improve the 3D Reconstruction of a Deployed Stent Michael Keyser (University of Utah) and Lucas Timmins (Biomedical Engineering, University of Utah) Faculty Mentor: Lucas Timmins (Biomedical Engineering, University of Utah) Introduction Coronary Heart Disease (CHD) is one of the most common causes of death in the United States. It accounts for 370,000 deaths annually. Various lifestyle choices, such as maintaining a healthy diet, exercising, not smoking, and limiting alcohol consumption to moderate amounts reduce the risk of CHD, but sometimes proper lifestyle choices are not sufficient [1]. CHD is caused by a buildup of plaque in the arteries called atherosclerosis. The plaque blocks blood flow to the myocardial tissue, which can lead to a heart attack. Clinicians use stents to open arteries blocked by plaque to restore blood flow in the affected artery. Stenting arteries is generally successful, but there is up to a 30% chance the stent will fail due to the ingrowth of tissue that re- blocks the vessel – termed restenosis [2]. Biomechanics has been shown to play a key role in stent failure [3]. Currently, studies that investigate the biomechanics in stented regions are limited to general cases because they lack the ability to model the in vivo geometry of the stented region. Elliott et al. developed a technique that utilizes optical coherence tomography (OCT) imaging data to reconstruct the in vivo deployed stent and vessel geometry [4]. In summary, the stent struts are reliably identified in OCT images [5]. Through fusion with biplane angiography [6], the stent struts are matched to the shape of the vessel to create a sparse representation of the known locations of the stent after being deployed, herein OCT point cloud. The physical dimensions of the stent are known, so an accurate model of the stent can be constructed and fitted to the OCT point cloud, through a constrained deformation process called diffeomorphic mapping [4]. Once the diffeomorphic mapping is complete, an accurate geometry of the patient specific stent has been reconstructed and the biomechanics in the region can be studied. The accuracy of the OCT point cloud is integral to the process because the diffeomorphic mapping assumes the OCT point cloud is correct. A rotational shift was observed in the collected OCT images. The curvature of the vessel caused the drift in OCT images. This imaging artifact is referred to as circumferential drift. OCT images can be adjusted prior to strut identification to correct for the effects of circumferential drift. An experimental testbed with three channels of constant curvature were scanned and the circumferential drift was measured across the OCT stack. The circumferential drift was observed in the direction of the curve. The normal vector of a curve in the Frenet-Serret frame (TNB) points inwards on a curve. Finally, the Frenet-Serret frame (TNB frame) was
← Previous Chapter Next Chapter →