Implementation of an electromagnetic tracking system for accurate intrahepatic puncture needle guidance: accuracy results in an in vitro model.

RATIONALE AND OBJECTIVES: Electromagnetic tracking potentially may be used to guide percutaneous needle-based interventional procedures. The accuracy of electromagnetic guided-needle puncture procedures has not been specifically characterized. This article reports the functional accuracy of a needle guidance system featuring real-time tracking of respiratory-related target motion. MATERIALS AND METHODS: A needle puncture algorithm based on a "free-hand" needle puncture technique for percutaneous intrahepatic portocaval systemic shunt was employed. Preoperatively obtained computed tomographic images were displayed on a graphical user interface and registered with the electromagnetically tracked needle position. The system and procedure was tested on an abdominal torso phantom containing a liver model mounted on a motor-driven platform to simulate respiratory excursion. The liver model featured two hollow tubes to simulate intrahepatic vessels. Registration and respiratory motion tracking was performed using four skin fiducials and a needle fiducial within the liver. Success rates for 15 attempts at simultaneous puncture of the two "vessels" of different luminal diameters guided by the electromagnetic tracking system were recorded. RESULTS: Successful "vessel" puncture occurred in 0%, 33%, and 53% of attempts for 3-, 5-, and 7-mm diameter "vessels," respectively. Using a two-dimensional accuracy prediction analysis, predicted accuracy exceeded actual puncture accuracy by 25%-35% for all vessel diameters. Accuracy outcome improved when depth-only errors were omitted from the analysis. CONCLUSIONS: Actual puncture success rate approximates predicted rates for target vessels 5 mm in diameter or greater when depth errors are excluded. Greater accuracy for smaller diameter vessels would be desirable for implementation in a broader range of clinical applications.