Improved Image Guidance for Minimally Invasive Liver Therapy: Image Co-registration Techniques

Abstract

Open surgeries are fast being replaced by minimally invasive procedures and noninvasive procedures. Image guidance is an indispensible tool during these procedures. When using navigation systems in minimally invasive procedures, pre-operative images are frequently used for guidance. However, these images do not display changes occurring during the surgery. Therefore, intra-operative imaging is also used. By combining intra-operative and pre-operative information, the procedure may be improved. The registrations of the pre-operative images with the updated intraoperative images will allow better overview and availability of needed information, which will give more accurate tool guidance and targeting during surgery. This PhD thesis gives an overview of technologies used in laparoscopic surgery. Registration methods and validation studies were explored and investigated to be able to improve the outcome of the minimally invasive procedures. Cone beam CT was explored as an intra-operative imaging modality in a laparoscopic setting. This was done to evaluate the deformations introduced in the liver due to pneumoperitoneum and respiration. Shifts as large as 44.5mm were found in this study. 4D ultrasound can be used as an intra-operative imaging technique as well as for the registration of pre-operative images to patient anatomy using 4D ultrasound images as an aid. A validation study of a motion estimation method was carried out on the liver of healthy volunteers. We found that the method was capable of tracking structures with an accuracy of 1mm. We studied the effect of non-periodic motion of drift in the liver owing to peristalsis movement and gravity. Initial results confirm a drift of 5.95mm comparing inhalations and 8.09 mm comparing exhalations when acquiring data over a period of 30 minutes. The surgeons can benefit more from the information available in the pre-operative CT/MR images in combination with intra-operative imaging such as ultrasound with more accurate image fusion techniques together with tool navigation and visualization. The work performed in this thesis with further development has potential for improving surgical outcome of minimally invasive liver therapy and also other abdominal soft tissue therapeutic procedures.