MR Navigation framework for robot-assisted endovascular intervention

Endovascular diseases are the number one cause of death in the western world. From atherosclerosis to stroke, there are many forms of this disease. Over the years, minimal invasive surgery (MiS) has proven to be superior in many cases to traditional open surgery. For MiS, many robotic platforms have been created to increase the repeatability, accuracy and decrease patient recovery time. Many of these platforms have been proven successful under Xray fluoroscopy. This imaging method brings danger of ionizing radiation to the patient and the surgeon, thus new imaging methods are researched. In 2021, the Imperial College London developed a robotic platform(CathBot) that is safe to use duringMagnetic Resonance Imaging (MRi). In this thesis, the port of CathBot from X-ray to MRi is approached, and comparing the pro’s and cons’ of MRi, a framework is created that tries to connect the robot to the MRi device. An imaging strategy is devised to detect one instrument of minimal invasive endovascular intervention (EvI), the guidewire, during cannulation of the abdominal aorta. Using traditional image processing and a combination of low sensitivity matching and blob finding, we are able to detect on average 68% of specific markers on the guidewire, with a standard deviation of 0.74 millimeter, and a bias of 4.16 millimeter, and calculate the distance from the guidewire to the vessel wall. The imaging can be done on a single core in real-time, at 40 frames per second on a mid/high end workstation.