Application accuracy study of a semipermanent fiducial system for frameless stereotaxis.

The accuracy of a semipermanent fiducial marker system developed at Wayne State University in collaboration with Fisher-Leibinger (Freiburg, Germany) was compared with reference to a standard stereotactic frame (Zamorano-Dujovny Localizing Unit; Fisher-Leibinger). For each patient in our study, 10 semipermanent markers were placed on the skull through a small incision and a pilot hole drilled for the marker; five markers were used for registration, and five were used for comparison. Gadolinium-enhanced magnetic resonance imaging was performed, and, upon registration using both ring and fiducial markers, 184 random points were collected by infrared digitization. All three-dimensional measurements (x, y, z) were converted into distance values correlating each value to the origin by the formula dij = SQRT (xij2 + yij2 + zij2). The mean difference of fiducial coordinates vs. absolute image coordinates was 1.72 +/- 0.42 mm (P = .0001), implying no significant difference. The mean difference in dij of the stereotactic ring coordinates vs. the absolute image coordinates was 3.35 +/- 0.59 mm (P = .00011). The mean difference in the fiducial markers vs. the stereotactic ring coordinates was 2.95 +/- 0.45 mm (P = .0001). All tests were declared significant at alpha = .016. The combination of interactive guidance with semipermanent fiducial markers allows for accurate localization of intracranial targets (as accurate or even more accurate than the stereotactic frame). Semipermanent fiducial markers facilitate the procedure logistically, allow for staged procedures (i.e., at the skull base or in epilepsy), and provide access for combined supra- and infratentorial approaches. We believe that the semipermanent fiducial markers system might represent an important development leading toward widespread use of interactive image guidance in conventional neurosurgery.