Evaluation of a novel, patient-mounted system for CT-guided needle navigation-an ex vivo study

To describe the features of a novel patient-mounted system for CT-guided needle navigation, the Puncture Cube System (PCS), and to evaluate the accuracy and efficiency of the PCS by (a) applying numerical simulations and (b) by conducting punctures using the system in comparison to punctures using the free-hand method (FHM).

The PCS consists of a self-adhesive cube that is attached to the patient, with multiple through-holes in the upper and lower template plate and dedicated software that, using a computer vision algorithm, recognizes the cube in a planning scan. The target in the image dataset is connected by a line, here "virtual needle," which passes through the cube. For any chosen path of the virtual needle, the entry points for the needle into the cube are displayed by the software for the upper and lower template on-the-fly. The possible exactness of the system was investigated by using numerical simulations. Next, 72 punctures were performed by 6 interventionists using a phantom to compare for accuracy, time requirement, and number of CT scans for punctures with the system to the FHM ex vivo (phantom study).

The theoretical precision to arrive at targets increased with the distance of the target but remained low. The mean error for targets up to 20 cm below the lower plate was computed to be well below 0.5 mm, and the worst-case error stayed below 1.3 mm. Compared to a conventional free-hand procedure, the use of the navigation system resulted in a statistically significantly improved accuracy (3.4 mm ± 2.3 mm versus FHM 4.9 mm ± 3.2 mm) and overall lower intervention time (168 s ± 28.5 s versus FHM 200 s ± 44.8 s). Furthermore, the number of CT scans was reduced to 2.3 versus FHM 2.8).

The PCS is a promising technique to improve accuracy and reduce intervention time in CT-guided needle navigations compared to the FHM.