In vivo quantification of arterial deformation due to pulsatile and non-pulsatile forces: Implications for the design of stents and stent-grafts

As more innovative endovascular devices are developed, the knowledge of dynamic changes in the vascular system has become increasingly important in ensuring the safety and efficacy of these devices. Regulatory agencies require specific benchtop testing of devices in a manner reflecting the in vivo environment in terms of the duty cycle and number of cycles. The superficial femoral and iliac arteries are particularly susceptible to multi-modal deformations due to repetitive hip and knee flexion, which can occur, for example, during walking. The coronary arteries experience dynamic forces due to cardiac motion, and the aortic vessel motion can be induced by cardiac, respiratory, and musculoskeletal motion. Understanding these cyclic changes in arterial geometry is essential for understanding device fatigue. Therefore, more complete and practical methods are needed to quantify three-dimensional (3D) arterial deformation. There is also a need to make these metrics readily accessible to device manufacturers and clinicians.;This dissertation describes methods for quantifying in vivo 3D arterial deformation due to pulsatile and nonpulsatile forces. I developed 3D centerline-based quantification methods by utilizing a two-dimensional segmentation technique to calculate the consistent centroid of the cross-sectional vessel lumen, and developed an optimal Fourier smoothing technique to eliminate spurious irregularities of the centerline connecting the centroids. Longitudinal strain and novel metrics for axial twist and curvature change were utilized to characterize 3D deformations of the arteries. By utilizing 3D level set segmentation methods, I extended centerline-based methods to surface-based quantification methods that directly utilize 3D segmentation data. The 3D surface-based quantification methods included a centerline extraction algorithm based on a minimum cross-sectional area calculation, and a strain twisting calculation algorithm based on mesh straightening and ostium shape matching algorithms. For curvature, optimal window size and best-fit torus calculation algorithms were developed. These methods were applied to quantify deformations of the superficial femoral artery, abdominal aorta, and common iliac artery due to musculoskeletal motion and deformations of the coronary artery due to cardiac pulsatile motion.;Musculoskeletal motion, specifically moderate hip flexion and knee flexion (hip flexion angle: 39+/-6°, knee flexion angle: 86+/-6°) for seven male healthy volunteers (age: 56+/-5 years), induced significant shortening (average strain: -6.9+/-1.9%, P < 0.001), twisting (average twist rate: 0.17+/-0.11°/mm, P < 0.05), and bending deformations (change in maxima curvature: 0.041+/-0.022 mm-1, P < 0.05) of the SFA. The coregistered SFAs based on the femur geometry revealed that the proximal portion of the SFA around the profunda femoris artery moved more inferiorly than the relatively immobile distal part of the SFA during hip and knee flexion. In vivo SFA kinematics obtained in this study provides insight into the interaction between the musculoskeletal and vascular systems. Moreover, maximal hip flexion (hip flexion angle: 134.0+/-9.7°), caused significant in vivo morphologic changes of the common iliac arteries for seven healthy subjects (age: 34+/-11 years), resulting in shortening (5.2+/-4.6%, P < 0.05), twisting (0.45+/-0.27°/mm, P < 0.05), and bending (0.015+/-0.007 mm-1, P < 0.05) of the arteries. Abdominal aorta also exhibited statistically significant shortening (2.9+/-2.1%, P < 0.05) and twist rate (0.07+/-0.05°/mm, P < 0.05). From the iliac geometry coregistration, predominantly superior translation of the common iliac arteries was observed.;Cardiac pulsatile motion also affected vascular morphology. Ventricular contraction induced significant deformations of the left anterior descending coronary artery (LAD). The mean absolute value of the LAD strain was 3.5+/-3.3% (P < 0.05) and twist rate was 0.44+/-0.44°/mm (P < 0.05). The branch angle between the LAD and left circumflex (LCX) decreased by 14.7+/-9.7° (P < 0.05) and the curvature measured at the maxima exhibited 17.7+/-22.1% increase in curvature (diastole: 0.16+/-0.10 mm-1, systole: 0.18+/-0.13 mm-1, P < 0.05).;These illustrative applications show the significance of each deformation metric, revealing significant longitudinal strain, axial twist and curvature change in the vessels due to pulsatile and non-pulsatile forces. The proposed methods and obtained results may aid in designing preclinical tests aimed at replicating dynamic in vivo conditions in the arterial tree for the purpose of developing more durable endovascular devices including stents and stent grafts.