12 Intra-operative Measurement of Brain Deformation
313
12.3.3 Intra-operative Ultrasound
3D iUS is a promising technology for measuring and monitoring brain shift during
tumour resection. 3D ultrasound can be reconstructed from tracked 2D US as
described in [83]. Commercial navigation systems, such as the Brainlab Curve
[8], have begun to support the use of 3D iUS for direct navigation and for
export via OpenIGTLink [84] for research in brain shift compensation. Several
groups have used iUS to measure brain shift since 1997, including [15, 25, 26,
60, 85–88]. Reinertsen and her colleagues have used Doppler US to measure the
displacements of blood vessels in the brain and used these to model brain shift
[89, 90]. Mohammedi et al. combined vessel tracking using Doppler US with
stereo images of the cortical surface to provide both surface and interior boundary
conditions to drive a patient-specific finite element model of the brain [91]. Morin
et al. combined vessel tracking using Doppler US with B-mode US of the cortical
surface [92]. Finally, transcranial ultrasound has been investigated as a noninvasive
way to track brain shift without interfering with the surgical field [93]. However,
many neurosurgeons are not trained to interpret US, and there are several practical
challenges when using transcranial US including ensuring a sterile field, requiring
a craniotomy large enough for the US probe, maintaining good fluid contact, and
deformation of the brain surface by the US probe which can impact deformation
measurements. Recent work has proposed a method to correct for surface contact
deformation in iUS [94, 95].
12.4 Conclusion
In summary, brain shift invalidates the assumption of most neuronavigation systems
that pre-operative image data can be rigidly registered to the patient during surgery.
Brain shift can be as large as 25 mm at the cortical surface and is often greater
than 3 mm at the deep tumour margin, where precision is required to achieve good
outcomes. Thus, it is critical that brain shift be accounted for in neuronavigation.
However, brain shift is complex, patient, and tumour specific, varies throughout
the brain, changes throughout surgery, and depends on many variables that are
hard to predict and model pre-operatively. For these reasons, the need to measure
and compensate for brain shift has been recognised by neurosurgeons and the
neuronavigation industry. Until recently, interventional and intra-operative MRI
has been the gold standard for handling brain shift. However, these systems are
expensive and disruptive to surgery, require a special surgical environment, and are
not available at most neurosurgical centres. Fortunately, several new technologies
have emerged for tracking brain shift both at the brain surface and deep structures.
We expect these technologies to lead to affordable neuronavigation systems that can
compensate for brain shift without disrupting the surgical workflow.
313
12.3.3 Intra-operative Ultrasound
3D iUS is a promising technology for measuring and monitoring brain shift during
tumour resection. 3D ultrasound can be reconstructed from tracked 2D US as
described in [83]. Commercial navigation systems, such as the Brainlab Curve
[8], have begun to support the use of 3D iUS for direct navigation and for
export via OpenIGTLink [84] for research in brain shift compensation. Several
groups have used iUS to measure brain shift since 1997, including [15, 25, 26,
60, 85–88]. Reinertsen and her colleagues have used Doppler US to measure the
displacements of blood vessels in the brain and used these to model brain shift
[89, 90]. Mohammedi et al. combined vessel tracking using Doppler US with
stereo images of the cortical surface to provide both surface and interior boundary
conditions to drive a patient-specific finite element model of the brain [91]. Morin
et al. combined vessel tracking using Doppler US with B-mode US of the cortical
surface [92]. Finally, transcranial ultrasound has been investigated as a noninvasive
way to track brain shift without interfering with the surgical field [93]. However,
many neurosurgeons are not trained to interpret US, and there are several practical
challenges when using transcranial US including ensuring a sterile field, requiring
a craniotomy large enough for the US probe, maintaining good fluid contact, and
deformation of the brain surface by the US probe which can impact deformation
measurements. Recent work has proposed a method to correct for surface contact
deformation in iUS [94, 95].
12.4 Conclusion
In summary, brain shift invalidates the assumption of most neuronavigation systems
that pre-operative image data can be rigidly registered to the patient during surgery.
Brain shift can be as large as 25 mm at the cortical surface and is often greater
than 3 mm at the deep tumour margin, where precision is required to achieve good
outcomes. Thus, it is critical that brain shift be accounted for in neuronavigation.
However, brain shift is complex, patient, and tumour specific, varies throughout
the brain, changes throughout surgery, and depends on many variables that are
hard to predict and model pre-operatively. For these reasons, the need to measure
and compensate for brain shift has been recognised by neurosurgeons and the
neuronavigation industry. Until recently, interventional and intra-operative MRI
has been the gold standard for handling brain shift. However, these systems are
expensive and disruptive to surgery, require a special surgical environment, and are
not available at most neurosurgical centres. Fortunately, several new technologies
have emerged for tracking brain shift both at the brain surface and deep structures.
We expect these technologies to lead to affordable neuronavigation systems that can
compensate for brain shift without disrupting the surgical workflow.