Open Access
Add to BibSonomyAbstract
Gas assisted laser machining of materials is a common practice in the manufacturing industry. Advantages in using gas assistance include reducing the likelihood of flare-ups in flammable materials and clearing away ablated material in the cutting path. Current surgical procedures and research do not take advantage of this and in the case for resecting osseous tissue, gas assisted ablation can help minimize charring and clear away debris from the surgical site. In the context of neurosurgery, the objective is to cut through osseous tissue without damaging the underlying neural structures. Different inert gas flow rates used in laser machining could cause deformations in compliant materials. Complications may arise during surgical procedures if the dura and spinal cord are damaged by these deformations. We present preliminary spinal deformation findings for various gas flow rates by using optical coherence tomography to measure the depression depth at the site of gas delivery.
© 2014 Optical Society of America
1. Introduction
Gas assisted laser machining of hard materials is widely used in manufacturing. Gabzydl performed significant experimentation on various flow rates of assist gas jets collinear with the cutting beam to determine efficiency when cutting various hard materials [1]. Using a gas assisted cutting process involves having a gas jet blow over the kerf during cutting, and presents distinct advantages. The use of inert gases reduces the likelihood of flare-ups while cutting flammable material. An assist gas jet also allows for the continuous clearing of the kerf so that the laser light is free to interact with the bottom of the kerf rather than being absorbed by cut debris [2,3]. Inert gas flow also allows smoke and plasma plumes to be blown away, giving the beam an obstruction-free path to the ablation site.
Even though gas assisted laser cutting is well understood in manufacturing and processing of hard materials, this is not the case in the context of surgery. Current surgical procedures that make use of lasers do not take advantage of an assist gas. In the application of hard tissue ablation, one potential benefit for gas assistance would be the improvement of etch rate (depth of material removed per laser pulse). Boppart et al. [4] as well as Leung et al. [5] have cited that carbonization as well as recrystallization of minerals hinders controlled surgical ablation. Carbonization is a result of high temperature heating at the surface of the tissue (resulting in charring) and potentially maybe damaging to surrounding biological tissue. By reducing the amount of flare-ups and by giving the beam an obstruction-free path, temperature heating may also be minimized.
Direct exposure of soft tissue underneath bone to the gas jet could occur once the laser has fully penetrated the protective osseous tissue. This could lead to inadvertent soft tissue damage if the velocity of the gas exiting the nozzle is too high and the standoff distance of nozzle from the cutting site is too small; this is especially concerning with regard to neural tissue, given its relative fragility. The spinal cord in particular may be sensitive to deformation changes, such as compression and distraction. These changes may be manifested as alterations in neurophysiological signals traversing the spinal cord, such as somatosensory evoked potentials (SSEPs) and motor evoked potentials (MEPs), and may indicate impending neurological injury. Previous work has demonstrated that distraction of porcine spinal cord up to 3.6% of its length of the spinal cord can cause loss of motor-evoked potentials [6]. Morris et al. [7] demonstrated loss of SSEP with spinal cord compression using a balloon catheter, however neither the amount of pressure nor depth of spinal cord deformation was quantified. It seems plausible that small amounts of pressure induced by gas flow could induce changes in SSEPs and MEPs. The goal of any gas assisted laser osteotomy system for neurosurgical applications would be to minimize damage to underlying neural tissue during momentary mechanical disruption/deformation.
Numerous studies have investigated the mechanical properties of the dura and spinal cord [8–11], with many studies examining the spinal cord’s response to loading and impact during spinal cord injury (SCI) [10,12–15]. However, to the authors knowledge at this time, no research has been done in situ on spinal cord deformation due to gas flow effects. High velocity instantaneous air impact systems have been developed but are used for brain and spine concussion models [16]. Several non-contact air impact systems have been developed for the purposed in finding tissue elasticity [17–20] but were focused on other biological tissues.
The depth of spinal cord deflection can be recorded with optical coherence tomography (OCT). Optical coherence tomography is a high resolution (1 – 10 μm) imaging modality that uses low power, low coherence light to produce near histological images of tissue in real-time. OCT works by initially separating a beam of laser light into two separate beams of equal length: the reference and sample beams. The sample beam is directed at perpendicular incidence onto the tissue. The depth-dependent backscatter is collected by the system and interfered with the reference beam. A resultant fringe signal, which is a function of path-length difference between the two beams, is then used to create an image with an axial resolution in the range of 3-15μm and a depth penetration of approximately 2mm depending on tissue optical property. Several authors [8,9] have pointed out that carbonization and recrystallization of minerals during hard tissue ablation scatters and absorbs incident light, therefore decreasing imaging penetration. This is a limitation for OCT imaging and can present shadowing artifacts that can appear in the images. Gas-assistance could potentially minimize carbonization and increase OCT imaging penetration to deliver better control in laser surgery.
In this initial study, OCT was used collinearly with a nitrogen jet to give real time deflection measurements of the spinal cord with respect to various volumetric flow rates. Depth profiles were distinct between spinal deformation with and without the intact dura. The depth profile can provide feedback information of mechanical stress for gas-assisted ablation on the spinal cord.
2. Methods and materials
2.1 Sample preparation
Two pig specimens were used for this study. A four-level laminectomy was performed on a fresh sacrificed adolescent pig where the spinal regions T14-L3 were exposed with the dura removed. Another four-level laminectomy was done on the second adolescent pig where L1-L4 was exposed with the dura intact. After collecting the spinal cord deformation data with the dura intact, the dura was then removed and the experiment repeated in the same locations.
2.2 In-line gas nozzle with OCT
Figure 1 shows a schematic diagram of the in-line gas nozzle coupled with a lens tube. The unit was coupled to the imaging system that consisted of a custom polygon-based swept source OCT (SS-OCT) with an A-scan sweeping frequency of 36 kHz. We refer the reader to [21] that describes the SS-OCT system in detail. Light from the sample arm is merged to the nozzle through a doublet collimator and focused through the gas-nozzle chamber. Similar systems such as Alonso-Canerio et al. [22] and Dorronsoro et al. [23] use a tonometer-based air-puff system for investigation into corneal dynamics. The tonometer systems that they used combined their OCT through a glass-window or tilted mirror. In contrast, our system combines the OCT gas-assisted nozzle as one integrated unit co-axially without additional optical elements in the custom designed nozzle chamber. The nitrogen source with a pressure meter delivers a known initial pressure to the system. A Victor Technologies nitrogen flow meter (NFM-TT) is connected near the inlet of the nozzle to measure the nozzle inlet flow. A vertical-mounted linear stage (Thorlabs) was used to make fine adjustments to precisely position the focal distance of the OCT. Nitrogen flows through the nozzle with an entrance diameter of 6.35mm, an exit diameter of 3mm, and a vertical height of 120mm. Total dura or spinal cord depression distance under various constant flow rates were recorded. Synchronization between the OCT system and starting flow rate was not required since the study was to show the maximum deflection of the spinal cord at any given flow rate. M-mode was used to record the depression profile at each specific location. We took the position of the M-mode image data and graphed it with respect to depth. The surface position was determined by running an edge detection algorithm over the M-mode images. The images were first background subtracted to enhance the surface feature; a median filter to remove speckle; and a threshold filter to identify the edge positions.
The separation distance between the nozzle exit and sample was approximately 2 mm. Two aiming lasers on the nozzle system converged to a focal point for OCT imaging. Saline was continuously used to keep the spinal tissue hydrated.
3. Results
3.1 Spinal deformation comparison with and without dura
A four-level laminectomy was performed on the first porcine specimen exposing T14 – L3. Data was collected at T14 and L1 locations as seen in Fig. 2. A volumetric flow rate reading of 60 standard cubic feet per hour (SCFH) in Fig. 3 shows the depression contour in spinal deformation for T14.
Spinal deformation data without the dura for T14 and L1 is shown in Fig. 4. Peak deflection for T14 was approximately 1 mm at 70 SCFH and 0.55 mm for L1. The relaxation time for spinal cord deformation recovery after the gas flow was turned off appears to increase with increasing gas flow. No spinal cord puncture was observed between the two areas for all flow rates.
A four-level laminectomy between L1 – L4 was performed on the second porcine specimen and is shown in Fig. 5. The deformation contour at L3 exposed to static 60 SCFH flow rate is shown in Fig. 6.
In situ spinal deformation results for various flow rates for L3 are shown in Fig. 7. No cases of dura or spinal cord puncture were observed throughout the period of study. The peak deformation increased with increasing flow rate. Peak displacement in L3 was considerably higher without the dura and had less mechanical oscillations due to gas flow disturbance. Deformation ranged between 0.02 mm to 0.5 mm with the dura intact and 0.02 mm to 0.9 mm without. Below SCFH 20, there was very little deformation change observed between the dura comparisons. A flow rate of 30 SCFH appears to be the threshold before there is significant change in deformation between the two cases. The relaxation period after discontinuation of gas flow was considerably steeper and shorter with dura intact in comparison to a smoother and longer curve without the dura. The relaxation time for spinal cord deformation recovery appears to be longer without the dura in comparison to the dura intact with increasing gas flow rates. The spinal cord appears to have undergone plastic deformation for both cases based on its final resting depression measurement.
In situ spinal deformation results for various flow rates for L4 are shown in Fig. 8. No punctures were observed for the spinal cord before and after the dura was removed. In comparison to L3, deformation in L4 showed little change between with and without dura cases. The peak deflection with dura for 70 SCFH was approximately 0.75 mm and 0.70 mm without the dura. In contrast to L3, the plastic recovery of the spinal cord (without the dura) in L4 after the gas was shut off did not appear to relax to its original baseline. The relaxation period for plastic recovery to baseline did not occur within the 35-second observation time. Future work will extend the temporal range to determine the elasticity of the spinal cord and its correlation to SSEP and MEP signals. This variation was due to lift from gas flowing under the spinal cord.
3.3 Spinal cord lift due to gas flow without the dura
Spinal cord lift was observed during two of the trials between L3 and L4 in which the gas flow created lift pressure under the spinal cord and distorted the deformation data. As observed in Fig. 4(a), 4(b) and 5(b), there is a clear separation in deformation above 30 SCFH. However, it can be seen in Fig. 8(b) that the deformation between flow rates 10 and 30-50 SCFH lie very close to one another. Figure 9 shows the effect due to spinal cord lift in comparison to Fig. 4(a), 4(b) and 5(b) where there is clean separation between the two different flow rates. Deformation offsets after subsequent gas flows in the same sampling location appear to contribute to zero displacement misalignment as seen in the figure. The relaxation level after the gas is turned off does not return to baseline but instead appears to settle between 0.1 – 0.2 mm.
4. Discussion
In this study, we examined the effects of spinal cord deformation from nitrogen gas flow of a porcine spinal cord. Using a custom-built gas nozzle in-line with the OCT, deformation images showing surface changes were taken in M-mode for various flow rates. These images were used to calculate differences in position between the initial state of the spinal cord at rest and the deflection due to constant, perpendicularly incident gas flow over the specimen.
The ability to use gas assistance in surgical laser applications offers several advantages that include reducing flare-ups and clearing away debris in the kerf during ablation. However, dura puncture can skew depth deformation measurements. It was shown in Fig. 8(b)) that the difference in spinal cord depression depth between 10 and 40 SCFH was negligible. Beyond 40 SCFH, the depression depth difference between the two cases was also negligible – the depression depth was approximately equal with and without the dura. It is possible that as nitrogen flowed on top of the spinal cord, circulating gas currents may have formed around and underneath the spinal cord in the epidural space, possibly creating lift, counteracting the depressive tendency of the incident gas flow. This is an important factor since lift causes additional vibration noise and obstructs the true depression depth in the area of interest. If the surgical events inadvertently were to cause a dural puncture near the site of the gas flow, dural inflation and/or debris being injected into the epidural space could be potential risks. Excess spinal cord lift height could be used to indicate to the surgeon that the dura has been compromised. A displacement difference of 0.4mm for L3 is shown in Fig. 10. This change may potentially have a significant impact on SSEP and MEP signals. In addition to potential SSEP and MEP signal changes, dural tearing results in a loss in cerebrospinal fluid (CSF) leakage. This can lead to neuroinflammation complications that may include postural headaches, vertigo, posterior neck pain, neck stiffness, nausea, diplopia, photophobia, tinnitus, and blurred vision [24,25]. In one study, it was found that patients with dural tearing had significantly higher hospital complications, higher mortality rates and higher total hospital charges [24]. Therefore, it would be beneficial for early detection of dural tearing during gas-assisted laser surgery that would help inform surgeons to prevent further impairment.
Understanding the impact of gas flow on spinal cord function is essential in assuring the integrity of the spinal cord during gas-assisted laser ablation procedures on the spine. Prior studies have examined various mechanically-induced models of spinal cord injury, however to date the authors have not identified specific studies involving gas-flow deformation techniques. In-vivo research in mechanically-induced spinal cord injury uses contact-based methods, distraction and shortening techniques, and spinal cord compression techniques. In-vivo contact-based impact systems strike the area of interest with a force that displaces the spinal cord and the physiological and mechanical characteristics are then measured [10–14]. Surgical distraction and shortening techniques involve changing the length of the spinal column, usually by osteotomy. Yang et al. [6] increased the length of the spinal cord in pigs by inserting stoppers at graded intervals, and examined physiological changes in spinal cord signals throughout the procedure to determine extent of spinal cord injury. Modi et al. [26] used staged shortening procedures via osteotomy to induce buckling in the spinal cord, and observed changes in MEPs and spinal cord blood flow throughout each stage. It is plausible, therefore, that changes in spinal cord function may be present during gas-assisted laser ablation procedures, particularly where spinal cord deformation may be observed.
Our initial study demonstrates the applicability in using gas-assistance for surgical ablation. Future work will include synchronization in starting and stopping of gas flows with OCT imaging. OCT imaging time will be increased to compensate for larger spinal cord deformations due to higher gas flows.
5. Conclusion
In conclusion, we have demonstrated an in-line SS-OCT integrated gas nozzle system for potential laser ablation applications. The dura forms a protective membrane that minimizes the gas flow deformation but can also increase mechanical noise in collecting depth data. More importantly, this technique can provide critical feedback during gas-assisted laser ablation in the proximity of neural tissue. This non-contact technique can also be extended to study the biomechanical properties of the spine cord or thecal sac in situ, during in-vivo studies, and future work could examine physiological changes in spinal cord function in concordance with this technique. To our knowledge, this is the first in situ demonstration on spinal cord deformation measurements using an in-line OCT gas nozzle integrated system.
Acknowledgments
The authors acknowledge assistance from Bruno Ponde and Minha Lee from Sunnybrook Health and Sciences Centre, Toronto, ON, Canada and financial support from Fed-Dev/Ontario Brain Institute. We would also like to thank the following people for their help in the study: Hamza Farooq, Helen Genis, Ryan Deorajh, and Atmiya Maisuria.
References and links
1. J. T. Gabzdyl, “The Effect of Laser Mode and Coaxial Gas Jet on Laser Cutting,” PhD thesis, University of London 1989.
2. R. S. Patel and M. Q. Brewster, “Gas-Assisted Laser-Metal Drilling: Experimental Results,” J. Thermophysics 5(1), 26–31 (1991). [CrossRef]
3. A. K. Dubey and V. Yadava, “Laser beam machining - A review,” Int. J. Mach. Tools Manuf. 48(6), 609–628 (2008). [CrossRef]
4. S. A. Boppart, J. Herrmann, C. Pitris, D. L. Stamper, M. E. Brezinski, and J. G. Fujimoto, “High-Resolution Optical Coherence Tomography-Guided Laser Ablation of Surgical Tissue,” J. Surg. Res. 82(2), 275–284 (1999). [CrossRef] [PubMed]
5. B. Y. C. Leung, P. J. L. Webster, J. M. Fraser, and V. X. D. Yang, “Real-Time Guidance of Thermal and Ultrashort Pulsed Laser Ablation in Hard Tissue Using Inline Coherent Imaging,” Lasers Surg. Med. 44(3), 249–256 (2012). [CrossRef] [PubMed]
6. J. H. Yang, S. W. Suh, H. N. Modi, E. T. Ramani, J. Y. Hong, J. H. Hwang, and W. Y. Jung, “Effects of vertebral column distraction on transcranial electrical stimulation-motor evoked potential and histology of the spinal cord in a porcine model,” J. Bone Joint Surg. Am. 95(9), 835–842 (2013). [CrossRef] [PubMed]
7. S. H. Morris, R. El-Hawary, J. J. Howard, and D. D. Rasmusson, “Validity of somatosensory evoked potentials as early indicators of neural compromise in rat model of spinal cord compression,” Clin. Neurophysiol. 124(5), 1031–1036 (2013). [CrossRef] [PubMed]
8. R. J. Fiford and L. E. Bilston, “The mechanical properties of rat spinal cord in vitro,” J. Biomech. 38(7), 1509–1515 (2005). [CrossRef] [PubMed]
9. P. Black, R. S. Markowitz, V. Cooper, A. Mechanic, H. Kushner, I. Damjanov, S. D. Finkelstein, and K. C. Wachs, “Models of Spinal Cord Injury: Part 1. Static Load Technique,” Neurosurgery 19(5), 752–762 (1986). [CrossRef] [PubMed]
10. N. R. Rodrigues, O. B. Letaif, A. F. Cristante, R. M. Marcon, R. P. Oliveira, and T. E. P. B. Filho, “Standardization of Spinal Cord Injury in Wistar Rats,” Acta Ortop. Bras. 18(4), 182–186 (2010).
11. J. T. Maikos, R. A. I. Elias, and D. I. Shreiber, “Mechanical Properties of Dura Mater from the Rat Brain and Spinal Cord,” J. Neurotrauma 25(1), 38–51 (2008). [CrossRef] [PubMed]
12. L. R. G. Vialle, L. R. Grochocki, and P. Nohama, “NYU device: automation of the Weight-drop rod,” IFMBE Proc. 14(2), 728–730 (2007). [CrossRef]
13. P. Black, R. S. Markowitz, V. Cooper, A. Mechanic, H. Kushner, I. Damjanov, S. D. Finkelstein, and K. C. Wachs, “Models of Spinal Cord Injury: Part 1. Static Load Technique,” Neurosurgery 19(5), 752–762 (1986). [CrossRef] [PubMed]
14. J. H. Kim, T. W. Tu, P. V. Bayly, and S. K. Song, “Impact Speed Does Not Determine Severity of Spinal Cord Injury in Mice with Fixed Impact Displacement,” J. Neurotrauma 26(8), 1395–1404 (2009). [CrossRef] [PubMed]
15. L. B. Jakeman, D. M. McTique, P. Walters, and B. T. Stokes, “The Ohio State University ESCID Spinal Cord Contusion Model,” Animal Models of Acute Neurological Injuries433–447 (2009).
16. Y. Xiong, A. Mahmood, and M. Chopp, “Animal models of traumatic brain injury,” Nat. Rev. Neurosci. 14(2), 128–142 (2013). [CrossRef] [PubMed]
17. S. Wang, K. V. Larin, J. Li, S. Vantipalli, R. K. Manapuram, S. Aglyamov, S. Emelianov, and M. D. Twa, “A focused air-pulse system for optical-coherence-tomography-based measurements of tissue elasticity,” Laser Phys. Lett. 10(7), 075605 (2013). [CrossRef]
18. S. Wang, A. L. Lopez 3rd, Y. Morikawa, G. Tao, J. Li, I. V. Larina, J. F. Martin, and K. V. Larin, “Noncontact quantitative biomechanical characterization of cardiac muscle using shear wave imaging optical coherence tomography,” Biomed. Opt. Express 5(7), 1980–1992 (2014). [CrossRef] [PubMed]
19. M. H. Lu, R. Mao, Y. Lu, Z. Liu, T. F. Wang, and S. P. Chen, “Quantitative Imaging of Young’s Modulus of Soft Tissues from Ultrasound Water Jet Indentation: A Finite Element Study,” Comput. Math. Methods Med. 2012, 979847 (2012). [CrossRef] [PubMed]
20. C. Cha and J. Oh, “An optofluidic mechanical system for elasticity measurement of thin biological tissues,” Biotechnol. Lett. 35(5), 825–830 (2013). [CrossRef] [PubMed]
21. C. Sun, B. Standish, B. Vuong, X. Y. Wen, and V. Yang, “Digital image correlation-based optical coherence elastography,” J. Biomed. Opt. 18(12), 121515 (2013). [CrossRef] [PubMed]
22. D. Alonso-Caneiro, K. Karnowski, B. J. Kaluzny, A. Kowalczyk, and M. Wojtkowski, “Assessment of corneal dynamics with high-speed swept source Optical Coherence Tomography combined with an air puff system,” Opt. Express 19(15), 14188–14199 (2011). [PubMed]
23. C. Dorronsoro, D. Pascual, P. Pérez-Merino, S. Kling, and S. Marcos, “Dynamic OCT measurement of corneal deformation by an air puff in normal and cross-linked corneas,” Biomed. Opt. Express 3(3), 473–487 (2012). [CrossRef] [PubMed]
24. H. Yoshihara and D. Yoneoka, “Incidental dural tear in spine surgery: analysis of a nationwide database,” Eur. Spine J. 23(2), 389–394 (2014). [CrossRef] [PubMed]
25. Y. Song, S. Li, B. Song, Y. Zhang, W. Gao, N. Li, K. Fan, and J. Ma, “The pathological changes in the spinal cord after dural tear with and without autologous fascia repair,” Eur. Spine J. 23(7), 1531–1540 (2014). [CrossRef] [PubMed]
26. H. N. Modi, S. W. Suh, J. Y. Hong, and J. H. Yang, “The effects of spinal cord injury induced by shortening on motor evoked potentials and spinal cord blood flow: an experimental study in swine,” J. Bone Joint Surg. Am. 93(19), 1781–1789 (2011). [CrossRef] [PubMed]
