Abstract
INTRODUCTION: To evaluate the accuracy of a quantitative 3D navigation system for CT-guided interventional procedures in a two-part study. MATERIALS AND METHODS: Twenty-two procedures were performed in abdominal and thoracic phantoms. Accuracies of the 3D anatomy map registration and navigation were evaluated. Time used for the navigated procedures was recorded. In the IRB approved clinical evaluation, 21 patients scheduled for CT-guided thoracic and hepatic biopsy and ablations were recruited. CT-guided procedures were performed without following the 3D navigation display. Accuracy of navigation as well as workflow fitness of the system was evaluated. RESULTS: In phantoms, the average 3D anatomy map registration error was 1.79 mm. The average navigated needle placement accuracy for one-pass and two-pass procedures, respectively, was 2.0 ± 0.7 mm and 2.8 ± 1.1 mm in the liver and 2.7 ± 1.7 mm and 3.0 ± 1.4 mm in the lung. The average accuracy of the 3D navigation system in human subjects was 4.6 mm ± 3.1 for all procedures. The system fits the existing workflow of CT-guided interventions with minimum impact. CONCLUSION: A 3D navigation system can be performed along the existing workflow and has the potential to navigate precision needle placement in CT-guided interventional procedures.