Ultrasound(US) is the cheapest, non-invasive and easily portable imaging modality compared to other conventional imaging modalities like MRI and CT. These advantages of US motivated us to replace C-ARM X-ray images with US images for navigation in a spinal surgery. However suboptimal quality of real-time US images and speckle noise patterns make bone detection in US images harder compared to X-Ray images. In order to achieve this goal, we propose a fast, image based bone segmentation and 2D-3D registration framework that operates with a tracked 2D US images and preoperative CT or MRI 3D volumes.
Our approach clearly reduces the intra-operative radiation exposure and huge space occupied by C-ARM inside the operating room.