Comparison between stereotactic CT and MRI coordinates of pallidal and thalamic targets using the Laitinen noninvasive stereoadapter.

The coordinates of one and the same target were compared between stereotactic CT and MRI studies, using the original Laitinen noninvasive Stereoadapter, and a slightly modified stereoadapter in 34 patients scheduled for pallidotomy or thalamotomy. The differences between CT and MRI coordinates were significant for the anteroposterior y (p < 0.001) and the vertical z (p < 0.01) coordinates. When the targets were analyzed separately for the coordinates in the right and left hemispheres, only those of the left-sided targets were significantly different between CT and MRI measurements. In patients where a vertex support was added to the Stereoadapter, there were no differences between CT and MRI target coordinates, regardless of the side of the target. However, in all patient groups, the three-dimensional vectorial difference between CT and MRI coordinates showed that the MRI-defined targets lay anterior and dorsal, that is, rostral, to the CT-defined targets, with a 95% confidence interval of the differences ranging from 1.8 to 2.4 mm. This rostral shift in target coordinates on MRI versus CT happens to coincide with the usual approach of the probe towards the target during surgery. It is concluded that the differences in target coordinates in our study are due partly to MRI distortion and partly to repositioning error of the Stereoadapter on the head. The relatively low magnitude of these differences does not preclude the use of the Stereoadapter for MRI-guided functional stereotactic surgery, provided careful impedance monitoring and macrostimulation of the target area prior to lesioning.