Automated Motion Control of the COAST Robotic Guidewire under Fluoroscopic Guidance

Peripheral arterial disease is one of the most prevalent cardiovascular diseases; its treatment is often catheter-based and requires the surgeon to manually navigate a guidewire to the affected region within the artery, usually with fluoroscopic images. It requires extensive skill and experience to navigate the guidewire to the target location and delays can cause increased radiation exposure to the surgeon. To overcome these challenges, we propose a fully automated approach to perform navigation of the COaxially Aligned STeerable (COAST) guidewire under fluoroscopic imaging in 2D phantom models. We utilize fluoroscopic images to calculate the optimal path between two points using a modified hybrid A-star algorithm in the phantom vasculature. The modified hybrid A-star computes a trajectory which is used for the velocity kinematics of the guidewire robot. The experiments show that the robot is able to follow the pre-computed path to the destination with a mean error of 8.2 pixels (2.87 mm).