CT-guided robotic placement of multiple convective delivery cannulas.

Abstract

Emerging data indicate that the use of cranial robotics with CT guidance can provide an efficient and effective alternative for the simultaneous placement of multiple convection-enhanced delivery (CED) infusion cannulas to targeted regions in the brain. To assess the feasibility, efficiency, and accuracy of robot-assisted, CT-guided infusion cannula placement, this study simultaneously placed multiple bilateral CED cannulas to clinically relevant targets in cranial phantoms. Simultaneous bilateral CT-guided robot-assisted infusion cannula placement was performed in 3D-printed phantom heads. Surgical and imaging (CT and MRI) results were evaluated. Five phantom heads were used in the study (30 targets, 6 targets per head). All cannulas were clearly identified on intraoperative CT and MRI. Targets were approached from frontal, parietal, and occipital entry points (10 targets each). The mean target depth was 72.9 (SD 15.1, range 50.2-97.8) mm. Based on CT imaging, the mean coronal error was 0.7 (SD 0.5, range 0.1-1.8) mm, sagittal error was 1.0 (SD 0.7, range 0.0-2.7) mm, and axial error was 0.6 (SD 0.5, range 0.1-1.8) mm. Based on MRI, the mean coronal error was 0.8 (SD 0.5, range 0.1-1.9) mm, sagittal error was 1.3 (SD 1.1, range 0.0-5.2) mm, and axial error was 0.9 (SD 0.7, range 0.1-2.7) mm. The mean radial error was similar comparing CT (0.4 [SD 0.7], range 0.01-2.3 mm) to MRI (0.6 [SD 0.7], range 0.02-2.9 mm). The time from initial target acquisition to placement of all cannulas was less than 30 minutes per phantom. Simultaneous CT-guided robotic insertion of multiple infusion cannulas was feasible, efficient, and precise. This technique could improve surgical efficiency for procedures involving targeted deliveries.