École Nationale Supérieure de
Techniques Avancées

Abstract

This study aims at providing complementary tools to prevent cardiovascular events. The goal is to characterize an early and patient-specific state of cardiovascular risk, by an analysis of carotid wall kinematics. Methods This study uses echographic videos and sets of medical data. It is based on a biomechanical model of elastic arteries, image processing (vessel segmentation and motion tracking), and object-oriented data processing. Results The group of people can be separated according to a precise score, or according to a profile. Both are representative of a given cardiovascular risk. Two kinematics come out as distinguishable, and characterize one mechanical state reflecting medical condition : radial kinematics and longitudinal kinematics. The maximum thickness of the vessel inner layers is 684 µm, and the mean of the radial movement amplitude is of 12,1%. Furthermore, it has been found that a state of patient-specific cardiovascular risk is not independent from given longitudinal kinematics. Differenciating typical kinematics (and thus, associated states of risk) is achieved by differenciating shapes of longitudinal motion. Interpretations are offered. They can be confirmed by a further improvement of the implemented image processings. Conclusion Results encourage a temporal monitoring of wall motions, to differentiate populations in terms of cardiovascular risk. THe amound of exploited data does not allow to conclude in a statistically significant way about quantitative values. Nevertheless, the study offers methods to continue to work on and resolve the objective.