Accuracy and feasibility of frameless stereotactic and robot-assisted CT-based puncture in interventional radiology:...

To compare the accuracy of frameless stereotactic and robot-assisted puncture in vitro based on computed tomography (CT) imaging with a slice thickness of 1, 3, and 5 mm.

300 punctures were carried out with help of the Atlas aiming device guided by the optical navigation system Stealth Station TREONplus and 150 punctures were guided by the robotic assistance system Innomotion. Conically shaped rods were punctured with Kirschner wires. The accuracy was evaluated on the basis of control CTs by measuring the Euclidean distance between the wire tip and target and the normal distance between the target and wire.

With the Stealth Station a mean Euclidean distance of 1.94 +/- 0.912, 2.2 +/- 1.136, and 2.74 +/- 1.166 mm at a slice thickness of 1, 3 and 5 mm, respectively, was reached. The mean normal distance was 1.64 +/- 0.919, 1.84 +/- 1.189, and 2.48 +/- 1.196 mm, respectively. The Innomotion system resulted in a mean Euclidean distance of 1.69 +/- 0.772, 1.91 +/- 0.673, and 2.30 +/- 0.881 mm, respectively, while the mean normal distance was 1.42 +/- 0.78), 1.60 +/- 0.733, and 1.98 +/- 1.002 mm, respectively. A statistical significance between accuracies with both systems with 1 mm and 3 mm slices could not be detected (p > 0.05). At a slice thickness of 5 mm, the robot was significantly more accurate, but not as accurate as when using thinner slices (p < 0.05). The procedure time is longer for the Innomotion system ( 30 vs. 18 min), and the practicability is higher with the Stealth Station.

The systems yield comparable accuracy. A slice thickness of 3 mm is adequate. Application of both methods in patient treatment can be expected to be safe and reliable.