Computer Vision News - November 2020

Katharina Breininger is a PhD student in thePatternRecognition Labat Friedrich- Alexander-Universität Erlangen- Nürnberg, under the supervision of Andreas Maier. Not only did she have the honour of presenting a paper at MICCAI 2020 last month, but her work was deemed to be of such high quality that she was awarded the title of Young Scientist. We are very grateful that she is taking some time out of her vacation to speak to us about her award-winning work and future plans. Endovascular aortic repair is a minimally invasive intervention that is used to treat aortic aneurysms . During this intervention there are stiff devices which are inserted to place stents and facilitate navigation. These devices deformthe vessels. A nephrotoxic iodine contrast agent is applied to visualize the vasculature and ensure navigation and stent placement is correct so that the intervention can be conducted safely. Being nephrotoxic, the use of the contrast agent must be limited as much as possible to minimize associated harm, such as kidney disease. Also, as it is an X-ray-guided procedure , exposure to X-ray is another risk which should be minimized, not only for patients but also for physicians. To solve this, this paper proposes to improve intraoperative image guidance by fusing preoperative information and intraoperative images , andmodeling the deformation seen intraoperatively that is not available in the preoperative images, when no devices have been inserted. Image fusion has only recently been established in clinical practice, but physicians are noticing its shortcomings when using only preoperative information. Katharina tells us her approach allows for correcting this deformation by smartly integrating user information. 2 Young Scientist Winner 14 Best of MICCAI 2020 Move over there: One-click deformation of preoperative 3D information to intraoperative X-ray during endovascular aortic repair