An image-guidance probe for stereotactic neurosurgery needs to address multiple challenges. The fundamental challenge is to provide information on what is in front of the probe in real time for a procedure that is otherwise blind. The primary clinical objective is to avoid lacerating blood vessels in front of the advancing probe. A secondary objective is to obtain information on the position of the probe tip relative to anatomic landmarks. This has to be accomplished with a thin needle-like geometry that will fit inside existing stereotactic apparatus.
The GRIN-rod-lens approach provides a solution to the multiple requirements for stereotactic neurosurgery. By using motion-free GRIN rod lenses as relay optics, the high speed A-line scanning rate of the laser system can be fully exploited without compromising the needle size. This high speed forward-imaging capability allows us to promptly avoid the at-risk vessels ( (
)) and locate the tip position ( (
)). Moreover, stable and high speed A-line scanning is critical for optimizing the flow VDR [37
]. Unstable and slow A-line scanning rates will degrade VDR and hinder vessel detection by DOCT. Our system not only detects vessels but also monitors the physiological changes of the blood flow ( (
)). Since there is no complex actuation at the distal end, this probe is one of the thinnest forward-imaging OCT probes. The diameter of the probe is a compromise between providing sufficient field of view and minimizing tissue injury. A GRIN lens diameter of 0.5 mm and field of view of 0.44 mm allows easy recognition of vessels, and the overall diameter of 0.74 mm is smaller than the inner diameter of existing stereotactic cannulae. Our current GRIN rod lens has a length of 12 cm. A portion of the rod lens (2.5 cm) is used for interfacing the rod lens with the scanning head. Therefore we can insert 9.5 cm deep inside the brain. The current probe length is more than adequate to reach all existing targets of DBS in the human. It can also be used for brain tumor biopsy, ventricle identification, etc. Also, the length could be further extended by splicing several pieces of GRIN rods together.
Evaluating the system in a living large mammalian brain provides correlates to the conditions of clinical stereotaxis. While useful information can be obtained from studies in phantoms, rodents and ex vivo
human tissue, these substitute samples cannot provide adequate simulation of the live human brain surgery. There are no vessels in the rat brain that are comparable in size to those in the human brain. Large animal brain imaging in vivo
also provided an opportunity to determine the extent of a potential problem for DOCT. The brain moves in conjunction with respiration and heartbeat. We found that this was problematic for imaging surface vessels, but not for intraparenchymal vessels. The surgeons can bypass surface vessels because they can directly visualize them. Once the probe entered the brain parenchyma, the movement problem is greatly decreased, possibly because the probe appears to stabilize the tissue directly in front of it. Additionally, vessels within the parenchyma can be easily pushed to the side by the probe ( (
)), whereas vessels in deep sulci ( (
)) that are tightly attached to the pia mater are not easily pushed aside by the probe. Vessels in sulci are really surface vessels that are buried by the cortical folds. This is consistent with anecdotal experience of neurosurgeons who meticulously avoid entering cortical sulci.
Besides detecting the vessels, the probe can also monitor the pulsation () and differentiate the vessel type (). These capabilities could be valuable for screening the vessels posing high risk in neurosurgery. We acknowledge that the aliasing problem may hinder using DOCT signal to quantify blood flow; however, we can work around this problem by using velocity variance [37
] or axial Kasai algorithm [40
]. Also, the high speed Fourier domain mode locking laser should be able to increase the velocity detection limit from one to two orders [41
In addition to detecting the at-risk vessels, this forward-imaging needle-type OCT probe can be an important complementary technology for current stereotaxic neurosurgery. Prior studies have demonstrated the ability of catheter-based OCT to provide information on the position of the probe relative to neighboring anatomic landmarks [26
]. Although the forward-scanning probe does not provide as much field of view as with the rotating side-imaging approach [26
], the needle probe can be used to provide real-time feedback on the degree to which the brain has shifted due to the surgery induced CSF leak. This is accomplished by establishing the probe tip relative to critical landmarks such as a prominent gray matter-white matter junction ( and
). Since the stereotactic system provides the expected position, any deviation between the expected and actual location of a landmark provides the surgeon useful information on the degree to which the brain has shifted during surgery. Secondly, it can often provide optical signatures of specific anatomic structures (i.e. the pencils of fibers in the putamen). The use of image guidance probes in conjunction with pre-operative MRI or CT images will give the surgeons greater confidence on the probe location. However, for diagnosis of other pathologies, a wider FOV may be desired. A possible way is to reconstruct a wider FOV image using manual scanning and optical tracking [43
]. The next phase of this project will be to miniaturize the overall proximal size and integrate a functional imaging modality. Besides the application in stereotactic procedures, this device can potentially be applied to many other image-guided interventions and for the detection of blood flow involving difficult to reach structures.