Frameless stereotaxy provides stereotactic localization without a stan
dard stereotactic frame. Other reference points on the patient are lin
ked to radiographic images by means of a three-dimensional (3D) digiti
zer. We report the use of a frameless stereotactic system to assist in
53 procedures in 52 children using the ISG Viewing Wand system (ISG T
echnologies, Missasauga, Ont.). This device uses surface landmarks as
well as random surface points as the reference system. The three-dimen
sional digitizer consists of a six-jointed articulated arm to which po
inters, a rigid ventriculoscope, or stereotactic biopsy probe guide ca
n be attached. Image reconstruction is performed on standard CT or MRI
(3D volume acquisition) images, with typically 3-mm slice thickness.
The tip of the probe can be displayed on 3D surfaced objects which can
be cut away to reveal reformatted gray scale displayed on the cut sur
faces, or in a triorthogonal view where simultaneous reformatted coron
al and sagital views at the level of the tip of the probe are displaye
d. Cases for which the viewing wand was used included craniotomies for
tumor (30), vascular malformation (6), epilepsy surgery (5), other le
sions (3), rigid ventriculoscopy (9), and stereotactic biopsy (2). The
accuracy of the system was measured by placing 20 fiducial markers on
a plastic head of which 3-mm thickness CT images had been made. The a
verage error using 4 fiducials as reference points was 1.5 mm when the
fiducials were distributed around the circumference of the head. Usin
g random surface points of 10, 20 and 40 in number gave average errors
of 2.3, 1.6, 1.5 mm, respectively. Twenty targets were implanted in a
cadaver head to which a BRW stereotactic frame (Radionics, Burlington
, Mass.) was attached and imaged in the same fashion. The average erro
r of the Viewing Wand, using 20-random surface point registration, was
2.3 mm, compared to 1.8 mm with the BRW frame using the are system. F
rameless stereotaxy is a valuable aid for a variety of neurosurgical p
rocedures in children providing rapid and relatively accurate stereota
ctic guidance without the hindrance of a stereotactic frame.