Registration of Intraoperative 3D Ultrasound with MR Data for the Navigated Computer Based Surgery [chapter]

Claudia Dekomien, Markus Mildenstein, Karin Hensel, Stephanie Hold, Susanne Winter
Springer Proceedings in Physics  
INTRODUCTION The trend towards minimally invasive surgery leads to an increasing demand for navigation systems. Image based navigation provides the surgeon a three dimensional orientation to control the position of the instruments within preoperative data of the patient. At the beginning of an image based navigated surgery the coordinate system of the preoperative image data needs to be registered within the coordinate system of the patient. Most of the commercial systems use landmarks as
more » ... landmarks as reference for the registration. The registration method developed by our group [1] is based on intraoperative freehand ultrasound imaging and overcomes some problems related to landmark based registration like invasiveness, inaccuracy and the time consumption of the process. Our system showed a high accuracy in phantom data [2] and reliable results in registering 3D ultrasound patient data with CT data of the lumbar spine [3]. Due to a good contrast in imaging of joint surrounding soft tissue and cartilage, MRI scans are of high diagnostic importance and often part of the preoperative routine. Hence navigation that is based on preoperative MRI images instead of CT data, is desirable especially for the knee surgery, where CT imaging plays a minor role. The main problem with MRI data compared to CT data is the difficulty of bone surface segmentation. Our method bases on a surface-volume registration algorithm which requires the extraction of the bone surface in the preoperative data. We started with a semi-automatic method, where we extracted that bone surface points from the MRI data that we can depict with a 3D ultrasound scan. In this work we show the first results of the registration of 3D ultrasound and MRI data of the femur and the tibia.
doi:10.1007/978-3-540-68764-1_42 fatcat:4yqypj5fyrenrklkgp4cbpowre