The relationship of imaging techniques to the accuracy of frameless stereotaxy

We analyzed the accuracy of a frameless stereotactic system using computed tomographic (CT) and magnetic resonance imaging (MRI) scans of different slice thickness and T(1) versus T(2) weighting of MRI.

An open skull with graphite pegs fixed to its base was used for all scans. CT scans were done with slice thicknesses of 1, 2 and 3 mm. MRI-visible markers were placed on top of pegs for T(1)-weighted and T(2)-weighted MRI scans, which were acquired at thicknesses of 1.5, 3 and 5 mm. For each scan, 3 separate registrations of a probe were performed; the distance between the actual probe location and that displayed on the registered image was noted. Each measurements was repeated 3 times for each registration.

Greatest accuracy was achieved with 3-mm-slice CT scans; this was not improved by using thinner slices. T(1)-weighted 1.5-mm MRI scans were 23% less accurate and T(2)-weighted 3-mm scans 37% less accurate.

Frameless stereotaxy should be done using CT scans when the greatest possible accuracy is desired. There appears to be no advantage to using slice thicknesses less than 3 mm. For most craniotomy applications, T(1)-weighted MRI using 3-mm slices provides sufficient accuracy. Lesions imaged only on T(2)-weighted MRI also can be approached with adequate precision using 3-mm scans.