Simulation of multi-curve active catheterization for endovascular navigation to complex targets.

The recent development of endovascular therapies has been accompanied by increasingly accurate navigation simulations to assist surgeons in decision making processes or to produce training tools. However, they have been focused mostly on targets within the aortic vasculature. In order to reach complex targets such as cerebral arteries by endovascular navigation, an active guidewire made of a Shape Memory Alloy (SMA) was recently proposed. The active part becomes deformed by the Joule effect and this deformation induces a bending of the guidewire. This setup is particularly suited for facilitating the access to Supra-Aortic Trunks (SATs) and, in our case, especially the left carotid artery. A complete characterization of the endovascular active navigation was conducted. In this framework, a test bench was developed to obtain an order of magnitude of the velocities applied on the guidewire as well as on the passive catheter going along with it in endovascular navigation. A numerical model was developed and validated in the case of navigation in a complex phantom aorta. We succeeded in representing crucial phenomena observed experimentally: snapping, active curvatures, interactions between the tools. In the last part of this study, it was demonstrated that adapting the guidewire design made it possible to hook the left carotid on three complex aortas.

Specific Ultrasound Data Acquisition for Tissue Motion and Strain Estimation: Initial Results.

Ultrasound applications such as elastography can benefit from 3-D data acquisition and processing. In this article, we describe a specific ultrasound probe, designed to acquire series of three adjacent imaging planes over time. This data acquisition makes it possible to consider the out-of-plane motion that can occur at the central plane during medium scanning, and is proposed with the aim of improving the results of strain imaging. In this first study, experiments were conducted on phantoms, and controlled axial and elevational displacements were applied to the probe using a motorized system. Radiofrequency ultrasound data were acquired at a 40-MHz sampling frequency with an Ultrasonix ultrasound scanner, and processed using a 3-D motion estimation method. For each of the 2-D regions of interest of the central plane in pre-compression data, a 3-D search was run to determine its corresponding version in post-compression data, with this search taking into account the region-of-interest deformation model chosen. The results obtained with the proposed ultrasound data acquisition and strain estimation were compared with results from a classic approach and illustrate the improvement produced by considering the medium's local displacements in elevation, with notably an increase in the mean correlation coefficients achieved.