Optimization and validation of a new 3D-US imaging robot to detect, localize and quantify lower limb arterial stenoses

Atherosclerosis is a disease caused by the accumulation of lipid deposits inducing the remodeling and hardening of the vessel wall, which leads to a progressive narrowing of arteries. These lesions are generally located on the coronary, carotid, aortic, renal, digestive and peripheral arteries. With regards to peripheral vessels, lower limb arteries are frequently affected. The severity of arterial lesions are evaluated by the stenosis degree (reduction > 50.0 % of the lumen diameter) using angiography, magnetic resonance angiography (MRA), computed tomography (CT) and ultrasound (US). However, to plan a surgical therapeutic intervention, a 3D arterial geometric representation is notably preferable. Imaging methods such as MRA and CT are very efficient to generate a three-dimensional imaging of good quality even though their use is expensive and invasive for patients.;The reliability of a 3D reconstructed geometry obtained with 2D-US images captured with a robotic system depends considerably on the positioning accuracy and the calibration procedure. Thus, the positioning accuracy of the robotic arm was evaluated in the workspace with a lower limb-mimicking phantom design (article 1 - chapter 3). In addition, a Z-phantom was designed to assure a precise calibration of the robotic system. These optimal methods were used to validate the system for the clinical application and to find the transformation which converts image coordinates of a 2D-ultrasound image into the robotic arm referential. From these results, all objects scanned by the robotic system can be adequately reconstructed in 3D.;Multimodal imaging vascular phantoms of lower limb arteries were used to evaluate the accuracy of the 3D representations (article 2 - chapter 4, article 3 - chapter 5). The validation of the reconstructed geometry with this system was performed by comparing surface points with the manufacturing vascular phantom file surface points. The accuracy to localize and quantify stenoses with the 3D-ultrasound robotic imaging system was also determined. These same evaluations were analyzed in vivo to perceive the feasibility of the study.;Keywords: 3D-US system, stenoses, calibration, medical robotics, calibration phantom, peripheral arterial diseases, vascular phantom, atherosclerosis.;3D-ultrasound can be perceived as a promising avenue in imaging for the location and the quantification of stenoses. This non invasive, non allergic (i.e., nephrotoxic contrast agent) and non-radioactive imaging modality offers distinct advantages in convenience, low cost and also multiple diagnostic options to quantify blood flow in Doppler. Since medical robots already have been used with success in surgery and orthopedics, our team has conceived a new medical 3D-US robotic imaging system to localize and quantify arterial stenoses in lower limb vessels. With this new technology, a clinician manually teaches the robotic arm the scanning path. Then, the robotic arm repeats with high precision the taught trajectory and controls simultaneously the ultrasound image acquisition process at even sampling and preserves safely the force applied by the US probe. Consequently, the reconstruction of a lower limb arterial geometry in 3D with this system could allow the location and quantification of stenoses with high accuracy. The objective of this research project consisted in validating and optimizing this 3D-ultrasound imaging robotic system.