306
S. Frisken et al.
12.2.1 Quantitative Results
Measurement of brain shift has typically been made either by observing displacements of landmarks on the cortical surface or by comparing pre-operative images
to images acquired during surgery or postoperatively. During tumour resection,
maximum displacements of up to 25 mm for points on the cortical surface have been
reported [17, 19, 23, 28]. Fahlbusch and Nimsky recorded brain shift at multiple
time points during surgery using an open magnet. With this frequent imaging, they
measured deformations that were larger than previously observed, with an average
maximum cortical displacement of 21.1 mm, 23.8 mm, and 37.6 mm for small,
medium, and large tumours, respectively [18]. At deep tumour margins, where
precision is critical, they observed brain shift greater than 3 mm in 66% of patients,
with low shifts (<2.9 mm) in 34% of cases, medium shifts (3–6.9 mm) in 42% of
cases, and high shifts (>7 mm) in 24% of cases [19].
During burr hole procedures for DBS, maximum displacements of 10–13 mm for
points on the cortical surface and smaller shifts of deep brain structures have been
observed [30, 32, 33, 35]. The shifts for deep brain structures were noted to be large
enough to compromise target location and to require multiple electrode adjustments.
Winkler et al. measured shifts of 2 mm at the subthalamic nucleus, a target of DBS
for Parkinson’s patients with typical dimensions 1.2 × 0.6 × 0.3 mm [30]. Ivan et
al. also found that 9% of patients who had burr hole surgery for electrode placement
had greater than 2 mm shifts at the target location [35].
In related studies, Schnaudigel et al. [36] found that changing the head orientation (left to right and supine to prone) in healthy subjects resulted in brain shifts of
almost 2 mm, and Faria et al. [37] measured up to 3 mm of brain deformation at the
cortical surface due to pulsatile motion from breathing and blood flow.
12.2.2 Qualitative Observations
Measurement of brain shift has shown that it is a complex phenomenon. It varies
by tumour location, head position, previous radiation treatment, tumour size, craniotomy size, and brain swelling [18]. Brain shift occurs both perpendicular to and
tangential to the craniotomy. There can be a significant difference between cortical
and subcortical brain shifts which researchers conclude cannot be predicted from
pre-operative imaging by biomechanical models [24] and particularly by models
that rely on gravity alone [21]. Significant deformation has been demonstrated
to occur even before resection begins [14, 27]. In addition, brain shift is time
dependent. It occurs throughout surgery and is nonuniform over time with large
shifts occurring when opening cysts, moderate shifts occurring during tumour
resection, and more gradual shifts occurring nearly continuously [20]. This work
suggests that biomechanical models of brain shift and neuronavigation systems that
handle brain shift will require intra-operative measurements to maintain accuracy.
S. Frisken et al.
12.2.1 Quantitative Results
Measurement of brain shift has typically been made either by observing displacements of landmarks on the cortical surface or by comparing pre-operative images
to images acquired during surgery or postoperatively. During tumour resection,
maximum displacements of up to 25 mm for points on the cortical surface have been
reported [17, 19, 23, 28]. Fahlbusch and Nimsky recorded brain shift at multiple
time points during surgery using an open magnet. With this frequent imaging, they
measured deformations that were larger than previously observed, with an average
maximum cortical displacement of 21.1 mm, 23.8 mm, and 37.6 mm for small,
medium, and large tumours, respectively [18]. At deep tumour margins, where
precision is critical, they observed brain shift greater than 3 mm in 66% of patients,
with low shifts (<2.9 mm) in 34% of cases, medium shifts (3–6.9 mm) in 42% of
cases, and high shifts (>7 mm) in 24% of cases [19].
During burr hole procedures for DBS, maximum displacements of 10–13 mm for
points on the cortical surface and smaller shifts of deep brain structures have been
observed [30, 32, 33, 35]. The shifts for deep brain structures were noted to be large
enough to compromise target location and to require multiple electrode adjustments.
Winkler et al. measured shifts of 2 mm at the subthalamic nucleus, a target of DBS
for Parkinson’s patients with typical dimensions 1.2 × 0.6 × 0.3 mm [30]. Ivan et
al. also found that 9% of patients who had burr hole surgery for electrode placement
had greater than 2 mm shifts at the target location [35].
In related studies, Schnaudigel et al. [36] found that changing the head orientation (left to right and supine to prone) in healthy subjects resulted in brain shifts of
almost 2 mm, and Faria et al. [37] measured up to 3 mm of brain deformation at the
cortical surface due to pulsatile motion from breathing and blood flow.
12.2.2 Qualitative Observations
Measurement of brain shift has shown that it is a complex phenomenon. It varies
by tumour location, head position, previous radiation treatment, tumour size, craniotomy size, and brain swelling [18]. Brain shift occurs both perpendicular to and
tangential to the craniotomy. There can be a significant difference between cortical
and subcortical brain shifts which researchers conclude cannot be predicted from
pre-operative imaging by biomechanical models [24] and particularly by models
that rely on gravity alone [21]. Significant deformation has been demonstrated
to occur even before resection begins [14, 27]. In addition, brain shift is time
dependent. It occurs throughout surgery and is nonuniform over time with large
shifts occurring when opening cysts, moderate shifts occurring during tumour
resection, and more gradual shifts occurring nearly continuously [20]. This work
suggests that biomechanical models of brain shift and neuronavigation systems that
handle brain shift will require intra-operative measurements to maintain accuracy.