Electrophysiological versus image-based targeting in the posteroventral pallidotomy

To study the functional accuracy of stereotactic targeting for the posteroventral pallidotomy (PVP), comparing targets chosen on magnetic resonance images (MRI), and fused MRI to computed tomographic (CT) images, with electrophysiologically refined anatomical targets. METHDOS AND MATERIALS: For each of the 10 pallidotomies three sets of targets were collected, beginning with the MRI targets. The second target set was measured on images generated by nonlinear volumetric fusion of MRI images with CT using Image Fusion (Radionics, Inc.). The anatomical target site was then determined electrophysiologically with intraoperative microelectrode recording and macroelectrode stimulation guidance.

Magnetic resonance imaging or MRI-CT fused images alone would not have been sufficiently accurate to preclude visual or motor complications in the posteroventral pallidotomy, based on our target located within 1 mm of the optic tract and within 2 mm of the internal capsule. In 2 of the 10 cases of either MRI or fused images, the targets were dangerously close to the optic tract. Two of 10 of the fused targets were within the internal capsule. The fusion of MRI with CT did not functionally improve the targeting accuracy of MRI, since the means of the MRI targets and the fused targets were statistically the same. Individually, however, the MRI target was different from the fused target in each case by an average radial distance of 3.5 +/- 2.3 mm, but such corrections were not statistically or surgically significant.

Image-based targeting including MRI or fused MRI-CT data may not be sufficiently accurate to prevent capsular or visual deficits in the posteroventral pallidotomy, necessitating electrophysiological refinement. In this report, the functional accuracy of MRI was not improved by fusion with CT.