Mechanical and pathophysiological in vivo characterization of the individual aortic wall based on 4D ultrasound imaging

This work focuses on the development of new non-invasive methods for the in vivo characteri-zation of the individual elastic behavior of the human aortic wall. Due to the physiological func-tion of the elastic properties within the cardiovascular system and its change with degenerative processes, this knowledge is of direct diagnostic relevance. In addition, the individual material properties are an important, prior to the work presented here unknown, determinant of patient-specific finite element models, which have been developed to calculate the maximum wall stress and to estimate the rupture risk of abdominal aortic aneurysms.
Using a modified device that provides time resolved 3D echocardiography with speckle tracking (4D ultrasound), a non-invasive in vivo full field measurement of human aortic wall motion was established in cooperation with industrial and clinical partners. It provided highly resolved mo-tion functions of discrete material points and derived in-plane strain tensors for wall surface segments with a sizes between 1 and about 20 mm^2. This new measurement of displacement and strain was validated in an in vitro experiment with respect to its agreement with optical full field measurements and to its reproducibility under identical conditions (test-retest reliability).
Two different methods for the analysis of aortic wall motion were developed and applied to different patient cohorts in clinical studies. A comparison of the cyclic three-dimensional de-formation of the proximal ascending and the abdominal aorta in terms of length and diameter change and twist provided deepened insight in the physiological “Windkessel” function of the ascending aorta. Statistical analysis of the distributions of local in-plane strains was used to ob-tain measures for size and heterogeneity of elastic aortic wall deformation (‘strain distribution indices’). A comparative clinical study in young volunteers and two groups of aged cardiovascular patients without and with abdominal aortic aneurysm showed that aortic walls reliably can be classified according to their cardiovascular health state by use of the obtained strain distribution indices and that these, therefore, are suited as new biomarker for cardiovascular health.
Two approaches were developed to characterize and model the individual elastic properties of the aortic wall. Firstly, a local distensibility coefficient has been introduced to linearly ap-proximate and identify the heterogeneous local functional elastic properties in the physiological range in vivo. Secondly, an iterative Finite Element Model Updating approach to the inverse identification of the individual orthotropic and hyperelastic constitutive behavior of geometrical-ly irregular aneurysmal walls was developed. It could be shown that constitutive parameter identification based on heterogeneous full field strain data is feasible even though only two load cases are accessible non-invasively in vivo. The approach was verified numerically and the effect of the measurement uncertainty on the constitutive parameter identification was examined. Finally, the approach was applied exemplarily to in vivo data of three patients of different age and cardiovascular health state.