gms | German Medical Science

Introduction: Laparoscopic surgery offers a lot of advantages [MW1] for the patient such as less trauma, faster postoperative mobilisation and less use of pain medication . However, drawbacks such as a narrow field of view, difficult spatial orientation and limited visualisation of structures at risk or target structures make this more complex minimally invasive therapies challenging for the surgeon and may lead to increased risk of complications. In laparoscopic colorectal surgery for patients having already had abdominal surgery injury of the ureter or major vessels may lead to prolonged operation time and conversion to open surgery along with long-time morbidity. Thus during surgery meticulous, time-consuming dissection is neccessary to identify and safely spare the ureter and vessels. To overcome these restraints a navigation system that employs preoperative CT-imaging for intraoperative guidance to identify target structures was developed.

Material and methods: The system consists of (1) a personal computer running the navigation software, (2) reflective spheres attached to the laparoscopic instruments that can be optically tracked with (3) an infrared-camera for optical tracking (NDI Polaris™) and (4) a vacuum mattress for immobilisation fixed on (5) a custom-made stretcher with attached markers for patient-image-registration. The study was performed with seven house pigs after approval by the local authorities.[MW1] After induction of total intravenous anaesthesia the pig was immobilised in the vacuum mattress. Subsequently contrast-enhanced CT-imaging of the pig was performed and the navigation system was installed in the operating room. Important target structures were segmented from the CT-images and 3D-models were generated. Pneumoperitoneum was applied and six ports were introduced. Next to the laparoscopic video a second screen showed a 3D-representation of the segmented organs. To evaluate the accuracy of the navigation system both ureters, aorta from left and right angle, both common iliac vessels, root of inferior mesenteric artery and the pelvic inlet were touched with a custom-made laparoscopic instrument that was tracked by the infrared-camera. Now the instrument position was tracked while touching the surface of the target structures. Afterwards the positioning error, i.e. the distance between computer-representation and measured position, was calculated. The results were analysed with descriptive statistics using the statistical computing environment R (http://www.r-project.org/).

Results: Of 56 possible target structures (7 pigs with 8 targets each) 42 could be measured (75%). A total of 2883 positions were measured. The overall target registration error was 6,8 mm (SD 4,4 mm) with 77% of the errors below 10 mm and 98,8% were below 20 mm. For the different target structures the error ranged from 5,3 mm (SD 3,2 mm) for the right ureter to 13,6 mm (SD 5,3 mm) for the root of the inferior mesenteric artery.

Conclusion: Navigation of retroperitoneal target structures was feasible with reasonable accuracy. The preoperative CT-imaging and intraoperative registration that were used might be sufficient to identify the target structures better and faster by limiting the search space, thus possibly decreasing operating time and the risk of complications as ureter or vessel injury. The next step is translation to clinical application.