- Email: [email protected]
Three-Dimensional Imaging in Neurosurgical Research and Education
Original Article
Three-Dimensional Imaging in Neurosurgical Research and Education Arnau Benet1-3, Halima Tabani1,2, Dylan Griswold1,2, Xin Zhang1,2, Olivia Kola1,2, Ali Tayebi Meybodi1,2, Michael T. Lawton1,2
OBJECTIVE: We describe the setup and use of different 3-dimensional (3-D) recording modalities (macroscopic, endoscopic, and microsurgical) in our laboratory and operating room and discuss their implications in neurosurgical research and didactics. We also highlight the utility of 3-D images in providing depth perception and discernment of structures compared with 2-dimensional (2-D) images.
residency programs may increase learning efficiency and shorten learning curves. However, use of 3-D imaging should not replace direct hands-on practice.
METHODS: The technical details for equipment and laboratory setup for obtaining 3-D images were described. The stereoscopic pair of images was obtained using a modified “shoot-shift-shoot” method and later converged to a 3-D image. For microsurgical procedures, 3-D images were obtained using an integrated 3-D video camera coupled to the surgical microscope in both the laboratory and the operating room. Illustrative cases were used to compare 2-D and 3-D images.
tereopsis, or depth perception, is the phenomenon whereby 2 slightly different images of an object perceived by each eye are superimposed by the brain to produce a 3-dimensional (3-D) image. Stereoscopic vision is an intrinsic part of our perception of the world and forms the basis for “prehension,” which is the ability to reach and grasp objects based on visual guidance.1 In this article, the terms “stereoscopic” and “3-dimensional” are used interchangeably. A cornerstone of surgical training is knowledge of the anatomy of the region of interest. Traditionally, anatomy was learned using illustrations depicted in textbooks and observation of procedures in the operating room (OR). Technologic advancement has ushered in a paradigm shift in the world of surgical didactics, with 3-D imaging emerging as a superior tool in neurosurgical education because it enables understanding of key spatial relationships between critical structures.2-5 Neurosurgical residency and fellowship training requires learning an increasing number of complex procedures in a short time. There is limited hands-on surgical experience during surgical procedures and via cadaveric workshops. Therefore, 3-D imaging for education and training has become invaluable in fields such as microneurosurgery and endoscopic skull base surgery, where the diminutive size of the structures and their close
-
-
RESULTS: Side-by-side comparisons of 2-D and 3-D images obtained using all modalities revealed that 3-D imaging was superior to 2-D imaging in providing depth perception and structure identification.
-
CONCLUSIONS: This is the first report in the literature of the methodology for obtaining 3-D endoscopic endonasal images using the 2-D endoscope. The use of 3-D imaging is invaluable in neurosurgical research and education, as it provides immediate depth perception (third dimension), allowing efficient understanding of key spatial relationships. Integration of 3-D imaging in neurosurgical
-
Key words Education - Endoscopic endonasal - Neurosurgery - Photography - Stereoscopic - Three-dimensional (3-D) -
Abbreviations and Acronyms 2-D: 2-Dimensional 3-D: 3-Dimensional CN: Cranial nerve OR: Operating room VA: Vertebral artery
WORLD NEUROSURGERY 91: 317-325, JULY 2016
INTRODUCTION
S
From the 1Department of Neurological Surgery, 2Skull Base and Cerebrovascular Laboratory, and 3Department of Otolaryngology Head and Neck Surgery, University of California, San Francisco, San Francisco, California, USA To whom correspondence should be addressed: Arnau Benet, M.D. [E-mail: [email protected]] Citation: World Neurosurg. (2016) 91:317-325. http://dx.doi.org/10.1016/j.wneu.2016.04.023 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2016 Elsevier Inc. All rights reserved.
www.WORLDNEUROSURGERY.org
317
ORIGINAL ARTICLE ARNAU BENET ET AL.
318
www.SCIENCEDIRECT.com
3-D IMAGING IN NEUROSURGERY
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2016.04.023
ORIGINAL ARTICLE ARNAU BENET ET AL.
relationships provide little margin of error. Studies have revealed that the use of 3-D imaging modalities in neurosurgical training is superior to conventional 2-dimensional (2-D) imaging in improving depth perception and anatomic understanding.2,6,7 Similarly, the use of 3-D endoscopes has been shown to boost performance, speed, and efficiency during neuroendoscopic procedures compared with traditional 2-D endoscopes.6-8 Although 3-D imaging has widespread implications in other fields (e.g., its use in the training of surgeons to perform laparoscopic procedures such as nephrectomy and cholecystectomy), the concept is still nascent in neurosurgery.9,10 Martins et al.5 and Shimizu et al.11 reported the procedure for obtaining and processing macroscopic 3-D images of cadaveric dissection specimens for use in neurosurgical education. However, reports are lacking on the setup and applications of 3-D recording modalities, including macroscopic, endoscopic, and microscopic, in the laboratory and in the neurosurgical OR as well as on the use of 3-D recording for neurosurgical education and research. Moreover, to our knowledge, there is no existing report of the procedure for obtaining 3-D images using currently available 2-D endoscopes. In this article, we describe the setup and procedure in the laboratory and OR at our institution for obtaining 3-D images using different modalities (macroscopic and endoscopic endonasal using surgical simulation in a cadaver and intraoperative microscopic using a surgical case illustration), and we discuss their implications in neurosurgical research and didactics. We also highlight the significance of 3-D imaging in providing anatomic orientation and depth perception by comparing 3-D images obtained using the aforementioned modalities with their 2-D counterparts. MATERIALS AND METHODS At the University of California, San Francisco, a fully integrated stereoscopic recording setup is available in the Skull Base and Cerebrovascular Laboratory and in the neurosurgical OR. In the laboratory, we use 3-D photography in conjunction with cadaveric surgical simulation to develop educational materials, including surgical anatomy lectures and instructional videos for medical students and neurosurgery residents.12 We also have a setup for endoscopic endonasal 3-D photography and microscopeintegrated 3-D microsurgical recording (TRENION 3D HD; Carl Zeiss, Jena, Germany; and TrueVision; TrueVision 3D Surgical, Santa Barbara, California, USA) using an operative microscope (OPMI PENTERO; Carl Zeiss, Jena, Germany). In the neurosurgical OR, we use integrated 3-D recording systems coupled to the operative microscope to project and record the operative procedures in 3-D images. We use live 3-D surgical broadcasting in the OR to guide the nursing team, surgical assistants, and
3-D IMAGING IN NEUROSURGERY
observers. We have also recorded 3-D videos for training and reporting complex cases as 3-D operative videos.13-18 The details of the setup and use of these 3 modalities are described next. 3-D Macroscopic Photography Equipment. The setup for high-resolution 3-D photography includes a reflex professional-grade 2-D digital camera (Nikon D810; Melville, NY, USA) coupled to either a macro lens (Nikkor 105 mm fixed zoom, 1:2.8 G; Melville, NY, USA) or a wide-range lens (Af-S Nikkor 24e120 mm, 1:4 G), mounted to a 3-D rig attachment (Redrock Micro Millimetered base and rotating camera shoes; El Dorado, CA, USA) on a tripod (Manfrotto model 055; Upper saddle river, NJ, USA). Three strobe lights (StrobeLite; Maumee, OH, USA) are used to provide lighting and reduce shadowing. Camera Settings. Before photography, the camera settings were adjusted according to the following protocol: shooting preset on aperture preference (F36 or maximum depth); format, raw nikon electronic format (NEF); image size, full frame (36 24 mm) and large resolution (7360 4912 pixels); focus, adjusted to the target of composition; flash, on; lighting, lens mounted flash (ring flash); white balance, flash light. Steps to Obtain Stereoscopic Pair of Images. We used a modified technique based on the “shoot-shift-shoot” method11 to obtain 2 stereoscopic images (mimicking images that would be viewed by each human eye). The tripod was set up at a distance of <1 m from the target. Care was taken to ensure that the tripod was at a horizontal level, parallel to the floor (using the leveler). The area of focus (target) was set at the center of the composition using the viewfinder of the camera. The 3-D photograph was obtained by taking a pair 2-D photographs, corresponding to the left and right eyes. The left photograph was arbitrarily obtained first by setting the convergence to 1.5 to the left; this was accomplished by rotating the camera onto the 3-D rig by 1.5 . The camera was then moved to the left through the 3-D rig until the center of the viewfinder was aligned with the target. If the target of interest was on a flat surface, the target locator in the viewfinder was adjusted to approximately 2 mm off to the same side (2 mm to the left of the dissection target for the left photograph and vice versa for the right photograph). The strobe lights coupled to accompanying light box fixtures were then oriented in the following manner: 2 strobe lights were set at each side of the camera, angled approximately 45 to point toward the target. The third strobe light was adjusted to lighten the shadows. The first 2-D photograph was taken to test all settings, and the light was adjusted to reduce shadows and obtain even lighting, if required. Once the optimal lighting was achieved, the final photograph was taken at these settings. Next, the convergence of the camera base
Figure 1. (A) Panoramic photograph and (B) floor plan of the operating room setup at our institution. The microscope is situated at the right side of the head of the patient, with the navigation unit located to the left of the surgeon. The 2dimensional (2-D) display, built in the body of the microscope, is directed toward the anesthesia team, which is positioned at the right of the patient’s torso. The 3-dimensional (3-D) display is set behind the surgeon, with ample space for residents and students to observe the procedure. There is an additional backup ceiling-mounted 2-D display monitor at the foot end of the patient pointed toward the nursing station.
WORLD NEUROSURGERY 91: 317-325, JULY 2016
www.WORLDNEUROSURGERY.org
319
ORIGINAL ARTICLE ARNAU BENET ET AL.
3-D IMAGING IN NEUROSURGERY
Figure 2. Macroscopic dissection using the retrosigmoid approach. The cerebellar tonsils, sigmoid sinus, vertebral artery, and cranial nerves VIIeXII were exposed using a retrosigmoid approach. The 2-dimensional (2-D) images are depicted on the left with their 3-dimensional (3-D) counterparts on the right. In the 2-D image, the vertebral artery appeared to be resting on the lateral mass of the C1 vertebra. However, in the 3-D image, the relative distance between the 2 structures was distinctly emphasized. Moreover, the spatial orientation of the different cranial nerves (VIIeXII) in relation to each other was better appreciated in the 3-D image compared with the 2-D image.
was adjusted to 1.5 to the right, or 3 from the original left position. The camera base was then slid horizontally along the 3-D rig until the target in the viewfinder matched the original target (the same anatomic point as that of the previous 2-D image). The second photograph was taken using the same light and identical camera setup. The stereoscopic pair was then processed for photographic optimization (color, exposure, highlights, shadows, and cropping) using dedicated software (Adobe Photoshop Lightroom version 5; Adobe Systems Incorporated, San Jose, California, USA). The stereoscopic pair of images was then set into a presentation software for display using methods similar to those reported by Martins et al.5 Endoscopic Endonasal 3-D Photography Endoscopic 3-D photographs were taken with a 2-D endoscope using 0 and 30 rigid scope attachments (Stryker Corporation, Kalamazoo, Michigan, USA). The 3-D photographs were obtained by taking a pair of 2-D photographs, no more than 5 mm apart. The procedure was started by aligning the dissection target to the center of the endoscope monitor. Next, the left photograph was taken by moving the endoscope in the horizontal plane 1e2.5 mm. The endoscope was then moved to the right, up to 5 mm from the left photograph position, and the corresponding right photograph was taken. The order of the photographs (3-D pair) was set arbitrarily. The distance between left and right photographs was set relative to the distance to the target: “the closer to the target, the lesser the distance.” Any movement along the vertical plane between the 2 photographs would result in an erroneous perspective of depth; therefore, vertical shift was explicitly avoided using anatomic landmarks as guides. 3-D Microsurgical (Deep Corridor) Recording Equipment. There are 2 widely used integrated systems for microsurgical 3-D recording available on the market (TRENION 3D HD and TrueVision). These systems universally comprise a built-in 3-D camera head, a signal cable, and a high-definition recorder. The built-in 3-D camera was attached to 1 of the optical outputs on the microscope head. The cable transmitting the
320
www.SCIENCEDIRECT.com
video signal connected the camera, which was coupled to the microscope, to the recording unit located nearby, separate from the microscope. This system is optimized for recording a 3-D video of a surgical procedure; 3-D still images were captured using the screenshot function. The system included a 3-D highdefinition display, through which live surgeries could be observed with 3-D glasses. At our institution, we have a dedicated intraoperative 3-D system set up in the OR for recording neurosurgical procedures (Figure 1). The 3-D photographs and videos obtained using the techniques explained in this article were used in educational meetings, resident lectures, and research publications.12-18 We used all modalities of 3-D photography acquisition described in this section (macroscopic, endoscopic, and microsurgical) to illustrate the difference in perception between 2-D and 3-D imaging in neurosurgery. We applied the principles of each technique to obtain significant illustrations of an endoscopic endonasal, macroscopic, and microsurgical procedure and displayed the 2-D and 3-D images side by side for comparison. Two cadavers were prepared for surgical simulation following our previously reported method.19 An endoscopic endonasal transclival approach was performed on 1 cadaver using the approach described before.20,21 The intradural stage of the approach was exposed, and 2-D and 3-D photographs were taken using the endoscopic method. In addition, a transcondylar and supracondylar far lateral approach was performed on another cadaver to expose the intradural stage, which was demonstrated in 2-D as well as 3-D images, using the macroscopic digital camera method. Finally, the virtual 3D library of the cerebrovascular service at the University of California, San Francisco, was searched to identify a 3-D operative video, and a case illustration was included to provide an example of the microsurgical 3-D recording method. RESULTS 3-D Macroscopic Photography of Transcranial Dissection The intradural stage of the transcondylar and supracondylar extensions of the far lateral approach was used for the
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2016.04.023
ORIGINAL ARTICLE ARNAU BENET ET AL.
photographic report (Figure 2). In the 2-D photograph (Figure 2 [left panel]), the vertebral artery (VA) appeared to be resting on the lateral mass of the C1. However, in the 3-D photograph (Figure 2 [right panel]), where the perspective of depth is present, the relative distance between the 2 structures was distinctly emphasized, showing the VA free from C1 and retracted off the surgical field. Moreover, the spatial orientation of the different cranial nerves (CN VIIeXII) in relation to each other and to the cerebellar hemisphere was better appreciated in the 3-D photograph. The depth from the condyle to CN XII, one of the most important concepts in efficient drilling during the transcondylar far lateral approach, was optimally appreciated in the 3-D photograph. The spatial relationship between CN XI and XII,
3-D IMAGING IN NEUROSURGERY
critical for drilling the jugular tubercle safely, was best understood in the 3-D photograph. Endoscopic Endonasal 3-D Photography The ventral compartment of the posterior fossa was exposed using an endoscopic endonasal transclival approach (Figure 3A and B). The 3-D photographs of the endoscopic endonasal transclival approach (Figure 3 [right panels]) provided a clear sense of depth of the surgical corridor compared with the 2-D photographs of the same exposure (Figure 3 [left panels]). The relative position of the pterygopalatine ganglion, the vidian nerve, and the internal carotid artery, critical for safety of the approach, could be understood efficiently in the 3-D
Figure 3. (A and B) Endoscopic endonasal dissection using the transclival approach. The 2-dimensional images are depicted on the left with their 3-dimensional (3-D) counterparts on the right. The different levels and depths at which the branches of the basilar artery originate (e.g., superior cerebellar artery, posterior cerebral artery, and anterior inferior cerebellar artery) were clearly appreciated in the 3-D image compared with the 2-dimensional image. The 3-D image highlighted the spatial relationship between the abducens nerve (cranial nerve VI) and the anterior inferior cerebellar artery, showing that cranial nerve VI traversed under the anterior inferior cerebellar artery near its origin at the pontomedullary sulcus. The relative position between the pterygopalatine ganglion and the internal carotid artery was also better understood in the 3-D image. AICA, anterior inferior cerebellar artery; BA, basilar artery; ICA, internal carotid artery (L, lacerum segment; PC, paraclival segment); IMA, internal maxillary artery; IOn, infraorbital nerve; PCA, posterior cerebral artery; Pg, pituitary gland; SCA, superior cerebellar artery; ON, optic nerve; VA, vertebral artery.
WORLD NEUROSURGERY 91: 317-325, JULY 2016
www.WORLDNEUROSURGERY.org
321
ORIGINAL ARTICLE ARNAU BENET ET AL.
322
www.SCIENCEDIRECT.com
3-D IMAGING IN NEUROSURGERY
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2016.04.023
ORIGINAL ARTICLE ARNAU BENET ET AL.
photograph (Figure 3B). Also, the depth and size of the paraclival segment of the internal carotid artery (Figure 3A) and its relationship in depth to CN VI (Figure 3B), which are very relevant to this step of the approach, were best shown in the 3-D photograph. The 3-D photograph in Figure 3A (right panel) provided a better understanding of the difference in depth between the superior cerebellar artery and the pituitary gland. It also clearly depicted how the CN VI traversed in relation to the anterior inferior cerebellar artery near its origin at the pontomedullary sulcus, a relationship not clearly appreciated in the 2-D photograph. Additionally, the relative distance between the anterior bend of the internal carotid artery and the posterior cerebral artery and CN III was best illustrated in the 3-D photograph (Figure 3B [right panel]). Microsurgical Case Illustration A 41-year-old woman presented with a recent history of imbalance on her right side, with multiple falls and garbled speech (Figure 4AeF). Angiography (Figure 4F [left panel]) revealed complete occlusion of both the proximal basilar artery and the left VA with a small right VA and absence of the left posterior communicating artery. An extended retrosigmoid approach for a V3 VA-to-anterior inferior cerebellar artery bypass with a saphenous vein graft was selected for blood flow augmentation in the posterior circulation (Figure 4AeE). Postoperative angiography (Figure 4F [right panel]) showed a patent bypass with robust restoration of blood flow. The patient had an uneventful postoperative course and remained symptom-free at postoperative 6-month follow-up. Figure 4AeE depicts 2-D intraoperative photographs in the left panel and the corresponding 3-D intraoperative photographs in the right panel. As evidenced in Figure 4, intraoperative 3-D photographs provided the opportunity for understanding the depth of the surgical corridor and allowed for the identification of relative distance and depth between CN XI and the anterior inferior cerebellar artery and the real angulation and position of the microinstruments, all of which were suboptimally shown through traditional 2-D imaging. The 3-D images allowed better appreciation of the positioning of the needle and the placement of suture throws on either side of the lumen of the interposition graft (Figure 4B and C), which was not adequately conveyed in the corresponding 2-D images, where the vessels appear flat. The 3-D images also provided better comprehension of the angulation and positioning of the clip relative to the surrounding structures. However, in terms of picture resolution, high-resolution 2-D and
3-D IMAGING IN NEUROSURGERY
3-D still photography using the macroscopic method, similar to that reported in Figure 2, were superior to the 3-D photographs obtained using the screenshot function during intraoperative microsurgical recording (Figure 4).
DISCUSSION Binocular vision is based on the principle that both eyes, situated at a small distance apart, view a target in a slightly different manner from each other. The 2 images, corresponding to each eye, are referred to as the “stereoscopic pair” and are superimposed to form a 3-D image, enabling depth perception. Depth perception is key in interpreting and understanding life for most of our daily tasks, but it becomes critical during surgery, as various structures are located in close proximity to each other, requiring skillful maneuvering of instruments to avoid unintentional damage to important neurovascular structures. This concept is best summarized by 2 axioms: “the better we see, the more we know,” and “the more we know, the more we see.”22 It is essential for the neurosurgical trainee to develop an excellent visual and tactile understanding of the surgical field. The trainee must learn to deftly navigate the surgical window to minimize complications and to ensure the patient’s optimal recovery and survival. Our side-by-side comparison of 2-D and 3-D recording (Figures 2e4) highlights the importance of 3-D imaging in providing depth perception and discerning closely related structures. Depth perception is important in all surgical procedures, but it takes on added importance when endoscopic endonasal techniques are undertaken, as demonstrated in our example using the transclival approach (Figure 3A and B). To our knowledge, this is the first report in the literature describing the method to obtain endoscopic endonasal 3-D images of neurosurgical procedures using currently available 2-D endoscopes. The surgical corridor of most endoscopic approaches is very deep and inaccessible to the naked eye, as it is situated beyond the nares. The relative sensation of depth of field, essential to endoscopic endonasal surgery, is felt indirectly by the surgeon, through dynamic movements of the endoscope and haptic feedback from accompanying instruments. This sensation of depth is unique to the surgeon and is very difficult to convey in 2D videos or photographs. Because 3-D vision bypasses the reliance on depth of field by touching, the trainee is no longer dependent on tactile sensation alone to learn, potentially shortening the learning curve, as validated previously in literature.8 Shah et al.7 reported that the use of the 3-D endoscope was preferred by novices compared with the 2-D endoscope and
Figure 4. (AeE) Intraoperative images for end-to-side anterior inferior cerebellar artery and V3 bypass using a saphenous vein graft. The 2-dimensional images are depicted on the left with their 3-D counterparts on the right. (F) Preoperative and postoperative angiography images are depicted in the left and right panels, respectively. Preoperative angiography revealed complete occlusion of the proximal basilar artery and the left vertebral artery with a small right vertebral artery. Postoperative angiography shows successful bypass with the restoration of blood flow; the star denotes the position of the graft. The 3-D image provided better appreciation of the relative distance and depth between cranial nerve XI and the anterior inferior cerebellar artery and the real angulation and position of the clip and microinstruments. In (B) and (C), compared with the 2-dimensional image, in which the vessel appeared flat, the 3-D image allowed better appreciation of the positioning of the needle and the placement of suture throws on either side of the lumen of the graft. AICA, anterior inferior cerebellar artery.
WORLD NEUROSURGERY 91: 317-325, JULY 2016
www.WORLDNEUROSURGERY.org
323
ORIGINAL ARTICLE ARNAU BENET ET AL.
3-D IMAGING IN NEUROSURGERY
helped improve depth perception, particularly during vascular dissection. Moreover, other studies have already revealed that 3-D neuroendoscopy is superior to 2-D endoscopy for the performance of various tasks, improving speed and efficiency and decreasing errors.6,8 However, despite the importance of 3-D recording for endoscopic endonasal procedures, use of 3-D endoscopic endonasal technology (Olympus, VisionSense, Endocam) is currently limited. This is due in part to decreased image resolution compared with 2-D endoscopic images, making it more difficult to discern important intraoperative structures and potentially leading to an increase in surgical complications. Furthermore, the sharp, quick, dynamic movements necessary for successful endonasal dissection may distort the 3-D experience, which works best with steady cameras or slow traveling; this might otherwise be described as intraoperative “motion sickness.” With ongoing advancements in endoscopic 3-D technology, we believe that these limitations can be overcome in the future, and the use of intraoperative 3-D endoscopic endonasal procedures is likely to become mainstream in teaching centers. Understanding the spatial relationships of structures is critical to deftly navigating the microsurgical corridor. Our case illustration using 3D intraoperative images for a VA-to-anterior inferior cerebellar artery bypass with a saphenous vein graft (Figure 4AeE) underscores the importance of 3-D imaging in understanding spatial relationships between different structures. The utility of 3-D intraoperative imaging in providing anatomic orientation, particularly to inexperienced trainees, has also been proven in previous reports. Rubino et al.2 studied the use of a 3-D intraoperative microscope during anterior circulation aneurysm surgery and compared it with traditional 2-D imaging among 30 physicians. They demonstrated that 3-D imaging was superior to the traditional 2-D still photographs in providing depth perception and improving the understanding of spatial relationships. Without depth perception provided by 3-D imaging, not only is the surgical trainee’s ability to process the visual field compromised, but also it becomes increasingly difficult for the instructor to properly emphasize the relative orientation of necessary surgical landmarks and key structures. Although cadaveric dissection is invaluable in gaining hands-on exposure to neurosurgical techniques, it does not prepare an inexperienced neurosurgical trainee to anticipate situations unique to surgery (e.g., bleeding that obscures the surgical field, brain expansion and volume changes, systolic-diastolic movements of the parenchyma).2 Observation of routine procedures using 3-D video enables trainees to observe the procedure from the eyes of the neurosurgeon, allowing for integration of maneuvers and the surgical technique from the unique perspective and view of the lead surgeon. Another advantage of 3-D video recording is that rare cases can be documented and used for research and teaching.23 Furthermore, in resource-poor areas where access to neurosurgical educators is limited, recorded videos can be used to introduce surgical trainees to techniques they would not otherwise be exposed to. With increasing opportunities to access and upload large amounts of data to cloud-based servers, the surgical classroom of the future may consist of a library of nearly every surgical approach recorded with 3-D imaging to be accessed by students worldwide. However, although 3-D
324
www.SCIENCEDIRECT.com
imaging helps increase familiarity with the anatomy, it is not a replacement for direct surgical exposure or hands-on surgical training. An important point underscored by our results is that in terms of picture resolution, 2-D still photography is superior to screenshots obtained from the built-in microsurgical 3-D video recording. This resolution gap may be explained by the size and density of the camera sensors and may be overcome as technology advances. However, there is no substitute for intraoperative 3-D recording, as the video playback allows the surgeon and trainee to move through the procedure in context, identifying key structures and their relationships as they are encountered during the procedure. Some surgeons may argue that the routine use of intraoperative 3-D microscopic documentation is costly and extravagant.24 We believe that incorporating a 3-D recording system in the OR is a wise investment, as it requires a one-time investment, with a large payback in educational opportunities, academic growth, and technical advancement. Teaching neurosurgical anatomy with 3-D imaging helps shorten learning curves, reducing the cost of training and improving surgical skill, which then improves patient outcomes and decreases overall health care costs. In their report, Barone et al.24 expressed the concern that the size of 3-D equipment may interfere with the routine functioning of the operating procedure. With the development of new technologies and incorporation of 3-D hardware and software in the intraoperative microscope, this limitation has been partly overcome. Our photograph of the OR setup at our institution and the floor-plan depict that with the use of state-of-the-art equipment with built-in 3-D cameras and wall mounted screens, the use of 3-D devices does not interfere with the activity in the OR. The wall-mounted 3-D display enables nurses and the surgical assistant to have a greater understanding of the surgical procedure and to anticipate surgical events and needs. Also, residents and students are able to observe the procedure from different standpoints in the OR without being in the way of the surgeon and nursing and anesthesia teams, as illustrated in Figure 1. In addition to the aforementioned practical applications of 3-D imaging, an emerging concept is computerized 3-D virtual reality simulation systems that render data from magnetic resonance imaging and computed tomography into virtual 3-D models. These systems are interactive and allow preoperative surgical simulation.25 Studies have shown that 3-D virtual reality systems are superior to traditional teaching methods (e.g., 2-D PowerPoint presentations) in teaching neuroanatomy to medical students.26 These systems have also been shown to be of value in training residents, allowing them to choose the appropriate approach specifically tailored to each patient.27 We suggest that 3-D images of anatomic dissections obtained using the methods we describe be used as an adjunct to enhance the virtual simulation experience, as these images provide a realistic depiction of the neurosurgical anatomy (e.g., brain dissection). CONCLUSIONS Use of 3-D imaging is invaluable in surgical anatomy research and education, as it enables novices and trainees to become oriented to neurosurgical anatomy and understand the key spatial
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2016.04.023
ORIGINAL ARTICLE ARNAU BENET ET AL.
3-D IMAGING IN NEUROSURGERY
relationships. 3-D recording should be the means by which the virtual classroom of surgical anatomy and procedures of the future is built, ensuring that medical students and surgical residents have access to the best and most relevant surgical education offered. Future advancements in this arena should focus on development of techniques for glasses-free viewing and improving the available 3-D virtual reality systems to set the stage
REFERENCES 1. Van Beurden MHPH, Van Hoey G, Hatzakis H, IJsselsteijn WA. Stereoscopic displays in medical domains: a review of perception and performance effects. In: Rogowitz BE, Pappas TN, eds. Proceedings of SPIE: Human Vision and Electronic Imaging XIV. Bellingham, WA: SPIE; 2009. 2. Rubino PA, Bottan JS, Houssay A, Salas López E, Bustamante J, Chiarullo M, et al. Three-dimensional imaging as a teaching method in anterior circulation aneurysm surgery. World Neurosurg. 2014;82:e467-e474. 3. Nakabayashi H, Shimizu K. Stereoscopic virtual realistic surgical simulation in intracranial aneurysms. Neurol India. 2012;60:191-197. 4. Christopher LA, William A, Cohen-Gadol AA. Future directions in 3-dimensional imaging and neurosurgery: stereoscopy and autostereoscopy. Neurosurgery. 2013;72(suppl 1):131-138. 5. Martins C, Ribas EC, Rhoton AL Jr, Ribas GC. Three-dimensional digital projection in neurosurgical education: technical note. J Neurosurg. 2015;123:1077-1080. 6. Fraser JF, Allen B, Anand VK, Schwartz TH. Three-dimensional neurostereoendoscopy: subjective and objective comparison to 2D. Minim Invasive Neurosurg. 2009;52:25-31. 7. Shah R, Leight W, Patel M, Surowitz JB, Wong YT, Wheless SA, et al. A controlled laboratory and clinical evaluation of a three-dimensional endoscope for endonasal sinus and skull base surgery. Am J Rhinol Allergy. 2011;25:141-144.
for the widespread use of 3-D imaging in surgical anatomy research and didactics. ACKNOWLEDGMENTS We express our gratitude to the body donors and their families, who through their altruism contributed to making this project possible.
ex-vivo laparoscopic tasks. J Endourol. 2012;26: 1085-1088. 11. Shimizu S, Tanaka R, Rhoton AL Jr, Fukushima Y, Osawa S, Kawashima M, et al. Anatomic dissection and classic three-dimensional documentation: a unit of education for neurosurgical anatomy revisited. Neurosurgery. 2006;58:E1000 [discussion: E1000]. 12. UCSF Skull Base & Cerebrovascular Laboratory. Available at: http://skullbaselab.ucsf.edu/. Accessed April 12, 2016.
approach to the clival and petroclival regions. Neurosurgery. 2009;65(6 suppl):42-50 [discussion: 50-52]. 21. Benet A, Prevedello DM, Carrau RL, RinconTorroella J, Fernandez-Miranda JC, Prats-Galino A, et al. Comparative analysis of the transcranial “far lateral” and endoscopic endonasal “far medial” approaches: surgical anatomy and clinical illustration. World Neurosurg. 2014;81:385-396. 22. Yas¸argil MG. Legacy of microneurosurgery: memoirs, lessons, and axioms. Neurosurgery. 1999; 45:1025-1092.
13. Tayebi Meybodi A, Benet A, Lawton MT. Microsurgical clipping of bilateral superior hypophyseal artery aneurysms through unilateral pterional craniotomy: 3-dimensional operative video [e-pub ahead of print]. Neurosurgery. http://dx.doi.org/10. 1227/NEU.0000000000001023.
23. Heath MD, Cohen-Gadol AA. Intraoperative stereoscopic 3D video imaging: pushing the boundaries of surgical visualisation and applications for neurosurgical education. Br J Neurosurg. 2012;26: 662-667.
14. Montemurro N, Benet A, Lawton MT. Clipping of ruptured residual anterior communicating artery aneurysm post endovascular coiling: 3dimensional operative video. Neurosurgery. 2016; 12:93-94.
24. Barone DG, Ban VS, Kirollos RW, Trivedi RA, Bulters DO, Ribas GC, et al. Alternative costeffective method to record 3D intra-operative images: a technical note. Br J Neurosurg. 2014;28: 819-820.
15. Benet A, Lawton MT. V3 vertebral artery to M2 middle cerebral artery bypass: 3-dimensional operative video [e-pub ahead of print]. Neurosurgery, 2015 Nov 4.
25. Bohm PE, Arnold PM. Simulation and resident education in spinal neurosurgery. Surg Neurol Int. 2015;6:33.
16. Benet A, Rincon-Torroella J, Lawton MT. Microsurgical clipping of a superior hypophyseal artery aneurysm: part 2: distal dural ring dissection and clipping: 3-dimensional operative video. Neurosurgery. 2015;11(suppl 2):359 [discussion: 359]. 17. Benet A, Rincon-Torroella J, Lawton MT. Microsurgical clipping of a superior hypophyseal artery aneurysm: part 1: intradural anterior clinoidectomy: 3-dimensional operative video. Neurosurgery. 2015;11(suppl 2):358 [discussion: 358].
8. Kawanishi Y, Fujimoto Y, Kumagai N, Takemura M, Nonaka M, Nakai E, et al. Evaluation of two- and three-dimensional visualization for endoscopic endonasal surgery using a novel stereoendoscopic system in a novice: a comparison on a dry laboratory model. Acta Neurochir (Wien). 2013;155:1621-1627.
18. Benet A, Lawton MT. Aneurysm transection and intra-aneurysmal clipping of a giant ophthalmic artery aneurysm: 3-dimensional operative video [e-pub ahead of print]. Neurosurgery, 2015 Jul 28.
9. Votanopoulos K, Brunicardi FC, Thornby J, Bellows CF. Impact of three-dimensional vision in laparoscopic training. World J Surg. 2008;32: 110-118.
19. Benet A, Rincon-Torroella J, Lawton MT, González Sánchez JJ. Novel embalming solution for neurosurgical simulation in cadavers. J Neurosurg. 2014;120:1229-1237.
10. Honeck P, Wendt-Nordahl G, Rassweiler J, Knoll T. Three-dimensional laparoscopic imaging improves surgical performance on standardized
20. de Notaris M, Cavallo LM, Prats-Galino A, Esposito I, Benet A, Poblete J, et al. Endoscopic endonasal transclival approach and retrosigmoid
26. Kockro RA, Amaxopoulou C, Killeen T, Wagner W, Reisch R, Schwandt E, et al. Stereoscopic neuroanatomy lectures using a threedimensional virtual reality environment. Ann Anat. 2015;201:91-98. 27. Agarwal N, Schmitt PJ, Sukul V, Prestigiacomo CJ. Surgical approaches to complex vascular lesions: the use of virtual reality and stereoscopic analysis as a tool for resident and student education. BMJ Case Rep. 2012;2012. Conflict of interest statement: The authors declare that the article content was composed in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Received 7 March 2016; accepted 6 April 2016
WORLD NEUROSURGERY 91: 317-325, JULY 2016
Citation: World Neurosurg. (2016) 91:317-325. http://dx.doi.org/10.1016/j.wneu.2016.04.023 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2016 Elsevier Inc. All rights reserved.
www.WORLDNEUROSURGERY.org
325
Three-Dimensional Imaging in Neurosurgical Research and Education Arnau Benet1-3, Halima Tabani1,2, Dylan Griswold1,2, Xin Zhang1,2, Olivia Kola1,2, Ali Tayebi Meybodi1,2, Michael T. Lawton1,2
OBJECTIVE: We describe the setup and use of different 3-dimensional (3-D) recording modalities (macroscopic, endoscopic, and microsurgical) in our laboratory and operating room and discuss their implications in neurosurgical research and didactics. We also highlight the utility of 3-D images in providing depth perception and discernment of structures compared with 2-dimensional (2-D) images.
residency programs may increase learning efficiency and shorten learning curves. However, use of 3-D imaging should not replace direct hands-on practice.
METHODS: The technical details for equipment and laboratory setup for obtaining 3-D images were described. The stereoscopic pair of images was obtained using a modified “shoot-shift-shoot” method and later converged to a 3-D image. For microsurgical procedures, 3-D images were obtained using an integrated 3-D video camera coupled to the surgical microscope in both the laboratory and the operating room. Illustrative cases were used to compare 2-D and 3-D images.
tereopsis, or depth perception, is the phenomenon whereby 2 slightly different images of an object perceived by each eye are superimposed by the brain to produce a 3-dimensional (3-D) image. Stereoscopic vision is an intrinsic part of our perception of the world and forms the basis for “prehension,” which is the ability to reach and grasp objects based on visual guidance.1 In this article, the terms “stereoscopic” and “3-dimensional” are used interchangeably. A cornerstone of surgical training is knowledge of the anatomy of the region of interest. Traditionally, anatomy was learned using illustrations depicted in textbooks and observation of procedures in the operating room (OR). Technologic advancement has ushered in a paradigm shift in the world of surgical didactics, with 3-D imaging emerging as a superior tool in neurosurgical education because it enables understanding of key spatial relationships between critical structures.2-5 Neurosurgical residency and fellowship training requires learning an increasing number of complex procedures in a short time. There is limited hands-on surgical experience during surgical procedures and via cadaveric workshops. Therefore, 3-D imaging for education and training has become invaluable in fields such as microneurosurgery and endoscopic skull base surgery, where the diminutive size of the structures and their close
-
-
RESULTS: Side-by-side comparisons of 2-D and 3-D images obtained using all modalities revealed that 3-D imaging was superior to 2-D imaging in providing depth perception and structure identification.
-
CONCLUSIONS: This is the first report in the literature of the methodology for obtaining 3-D endoscopic endonasal images using the 2-D endoscope. The use of 3-D imaging is invaluable in neurosurgical research and education, as it provides immediate depth perception (third dimension), allowing efficient understanding of key spatial relationships. Integration of 3-D imaging in neurosurgical
-
Key words Education - Endoscopic endonasal - Neurosurgery - Photography - Stereoscopic - Three-dimensional (3-D) -
Abbreviations and Acronyms 2-D: 2-Dimensional 3-D: 3-Dimensional CN: Cranial nerve OR: Operating room VA: Vertebral artery
WORLD NEUROSURGERY 91: 317-325, JULY 2016
INTRODUCTION
S
From the 1Department of Neurological Surgery, 2Skull Base and Cerebrovascular Laboratory, and 3Department of Otolaryngology Head and Neck Surgery, University of California, San Francisco, San Francisco, California, USA To whom correspondence should be addressed: Arnau Benet, M.D. [E-mail: [email protected]] Citation: World Neurosurg. (2016) 91:317-325. http://dx.doi.org/10.1016/j.wneu.2016.04.023 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2016 Elsevier Inc. All rights reserved.
www.WORLDNEUROSURGERY.org
317
ORIGINAL ARTICLE ARNAU BENET ET AL.
318
www.SCIENCEDIRECT.com
3-D IMAGING IN NEUROSURGERY
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2016.04.023
ORIGINAL ARTICLE ARNAU BENET ET AL.
relationships provide little margin of error. Studies have revealed that the use of 3-D imaging modalities in neurosurgical training is superior to conventional 2-dimensional (2-D) imaging in improving depth perception and anatomic understanding.2,6,7 Similarly, the use of 3-D endoscopes has been shown to boost performance, speed, and efficiency during neuroendoscopic procedures compared with traditional 2-D endoscopes.6-8 Although 3-D imaging has widespread implications in other fields (e.g., its use in the training of surgeons to perform laparoscopic procedures such as nephrectomy and cholecystectomy), the concept is still nascent in neurosurgery.9,10 Martins et al.5 and Shimizu et al.11 reported the procedure for obtaining and processing macroscopic 3-D images of cadaveric dissection specimens for use in neurosurgical education. However, reports are lacking on the setup and applications of 3-D recording modalities, including macroscopic, endoscopic, and microscopic, in the laboratory and in the neurosurgical OR as well as on the use of 3-D recording for neurosurgical education and research. Moreover, to our knowledge, there is no existing report of the procedure for obtaining 3-D images using currently available 2-D endoscopes. In this article, we describe the setup and procedure in the laboratory and OR at our institution for obtaining 3-D images using different modalities (macroscopic and endoscopic endonasal using surgical simulation in a cadaver and intraoperative microscopic using a surgical case illustration), and we discuss their implications in neurosurgical research and didactics. We also highlight the significance of 3-D imaging in providing anatomic orientation and depth perception by comparing 3-D images obtained using the aforementioned modalities with their 2-D counterparts. MATERIALS AND METHODS At the University of California, San Francisco, a fully integrated stereoscopic recording setup is available in the Skull Base and Cerebrovascular Laboratory and in the neurosurgical OR. In the laboratory, we use 3-D photography in conjunction with cadaveric surgical simulation to develop educational materials, including surgical anatomy lectures and instructional videos for medical students and neurosurgery residents.12 We also have a setup for endoscopic endonasal 3-D photography and microscopeintegrated 3-D microsurgical recording (TRENION 3D HD; Carl Zeiss, Jena, Germany; and TrueVision; TrueVision 3D Surgical, Santa Barbara, California, USA) using an operative microscope (OPMI PENTERO; Carl Zeiss, Jena, Germany). In the neurosurgical OR, we use integrated 3-D recording systems coupled to the operative microscope to project and record the operative procedures in 3-D images. We use live 3-D surgical broadcasting in the OR to guide the nursing team, surgical assistants, and
3-D IMAGING IN NEUROSURGERY
observers. We have also recorded 3-D videos for training and reporting complex cases as 3-D operative videos.13-18 The details of the setup and use of these 3 modalities are described next. 3-D Macroscopic Photography Equipment. The setup for high-resolution 3-D photography includes a reflex professional-grade 2-D digital camera (Nikon D810; Melville, NY, USA) coupled to either a macro lens (Nikkor 105 mm fixed zoom, 1:2.8 G; Melville, NY, USA) or a wide-range lens (Af-S Nikkor 24e120 mm, 1:4 G), mounted to a 3-D rig attachment (Redrock Micro Millimetered base and rotating camera shoes; El Dorado, CA, USA) on a tripod (Manfrotto model 055; Upper saddle river, NJ, USA). Three strobe lights (StrobeLite; Maumee, OH, USA) are used to provide lighting and reduce shadowing. Camera Settings. Before photography, the camera settings were adjusted according to the following protocol: shooting preset on aperture preference (F36 or maximum depth); format, raw nikon electronic format (NEF); image size, full frame (36 24 mm) and large resolution (7360 4912 pixels); focus, adjusted to the target of composition; flash, on; lighting, lens mounted flash (ring flash); white balance, flash light. Steps to Obtain Stereoscopic Pair of Images. We used a modified technique based on the “shoot-shift-shoot” method11 to obtain 2 stereoscopic images (mimicking images that would be viewed by each human eye). The tripod was set up at a distance of <1 m from the target. Care was taken to ensure that the tripod was at a horizontal level, parallel to the floor (using the leveler). The area of focus (target) was set at the center of the composition using the viewfinder of the camera. The 3-D photograph was obtained by taking a pair 2-D photographs, corresponding to the left and right eyes. The left photograph was arbitrarily obtained first by setting the convergence to 1.5 to the left; this was accomplished by rotating the camera onto the 3-D rig by 1.5 . The camera was then moved to the left through the 3-D rig until the center of the viewfinder was aligned with the target. If the target of interest was on a flat surface, the target locator in the viewfinder was adjusted to approximately 2 mm off to the same side (2 mm to the left of the dissection target for the left photograph and vice versa for the right photograph). The strobe lights coupled to accompanying light box fixtures were then oriented in the following manner: 2 strobe lights were set at each side of the camera, angled approximately 45 to point toward the target. The third strobe light was adjusted to lighten the shadows. The first 2-D photograph was taken to test all settings, and the light was adjusted to reduce shadows and obtain even lighting, if required. Once the optimal lighting was achieved, the final photograph was taken at these settings. Next, the convergence of the camera base
Figure 1. (A) Panoramic photograph and (B) floor plan of the operating room setup at our institution. The microscope is situated at the right side of the head of the patient, with the navigation unit located to the left of the surgeon. The 2dimensional (2-D) display, built in the body of the microscope, is directed toward the anesthesia team, which is positioned at the right of the patient’s torso. The 3-dimensional (3-D) display is set behind the surgeon, with ample space for residents and students to observe the procedure. There is an additional backup ceiling-mounted 2-D display monitor at the foot end of the patient pointed toward the nursing station.
WORLD NEUROSURGERY 91: 317-325, JULY 2016
www.WORLDNEUROSURGERY.org
319
ORIGINAL ARTICLE ARNAU BENET ET AL.
3-D IMAGING IN NEUROSURGERY
Figure 2. Macroscopic dissection using the retrosigmoid approach. The cerebellar tonsils, sigmoid sinus, vertebral artery, and cranial nerves VIIeXII were exposed using a retrosigmoid approach. The 2-dimensional (2-D) images are depicted on the left with their 3-dimensional (3-D) counterparts on the right. In the 2-D image, the vertebral artery appeared to be resting on the lateral mass of the C1 vertebra. However, in the 3-D image, the relative distance between the 2 structures was distinctly emphasized. Moreover, the spatial orientation of the different cranial nerves (VIIeXII) in relation to each other was better appreciated in the 3-D image compared with the 2-D image.
was adjusted to 1.5 to the right, or 3 from the original left position. The camera base was then slid horizontally along the 3-D rig until the target in the viewfinder matched the original target (the same anatomic point as that of the previous 2-D image). The second photograph was taken using the same light and identical camera setup. The stereoscopic pair was then processed for photographic optimization (color, exposure, highlights, shadows, and cropping) using dedicated software (Adobe Photoshop Lightroom version 5; Adobe Systems Incorporated, San Jose, California, USA). The stereoscopic pair of images was then set into a presentation software for display using methods similar to those reported by Martins et al.5 Endoscopic Endonasal 3-D Photography Endoscopic 3-D photographs were taken with a 2-D endoscope using 0 and 30 rigid scope attachments (Stryker Corporation, Kalamazoo, Michigan, USA). The 3-D photographs were obtained by taking a pair of 2-D photographs, no more than 5 mm apart. The procedure was started by aligning the dissection target to the center of the endoscope monitor. Next, the left photograph was taken by moving the endoscope in the horizontal plane 1e2.5 mm. The endoscope was then moved to the right, up to 5 mm from the left photograph position, and the corresponding right photograph was taken. The order of the photographs (3-D pair) was set arbitrarily. The distance between left and right photographs was set relative to the distance to the target: “the closer to the target, the lesser the distance.” Any movement along the vertical plane between the 2 photographs would result in an erroneous perspective of depth; therefore, vertical shift was explicitly avoided using anatomic landmarks as guides. 3-D Microsurgical (Deep Corridor) Recording Equipment. There are 2 widely used integrated systems for microsurgical 3-D recording available on the market (TRENION 3D HD and TrueVision). These systems universally comprise a built-in 3-D camera head, a signal cable, and a high-definition recorder. The built-in 3-D camera was attached to 1 of the optical outputs on the microscope head. The cable transmitting the
320
www.SCIENCEDIRECT.com
video signal connected the camera, which was coupled to the microscope, to the recording unit located nearby, separate from the microscope. This system is optimized for recording a 3-D video of a surgical procedure; 3-D still images were captured using the screenshot function. The system included a 3-D highdefinition display, through which live surgeries could be observed with 3-D glasses. At our institution, we have a dedicated intraoperative 3-D system set up in the OR for recording neurosurgical procedures (Figure 1). The 3-D photographs and videos obtained using the techniques explained in this article were used in educational meetings, resident lectures, and research publications.12-18 We used all modalities of 3-D photography acquisition described in this section (macroscopic, endoscopic, and microsurgical) to illustrate the difference in perception between 2-D and 3-D imaging in neurosurgery. We applied the principles of each technique to obtain significant illustrations of an endoscopic endonasal, macroscopic, and microsurgical procedure and displayed the 2-D and 3-D images side by side for comparison. Two cadavers were prepared for surgical simulation following our previously reported method.19 An endoscopic endonasal transclival approach was performed on 1 cadaver using the approach described before.20,21 The intradural stage of the approach was exposed, and 2-D and 3-D photographs were taken using the endoscopic method. In addition, a transcondylar and supracondylar far lateral approach was performed on another cadaver to expose the intradural stage, which was demonstrated in 2-D as well as 3-D images, using the macroscopic digital camera method. Finally, the virtual 3D library of the cerebrovascular service at the University of California, San Francisco, was searched to identify a 3-D operative video, and a case illustration was included to provide an example of the microsurgical 3-D recording method. RESULTS 3-D Macroscopic Photography of Transcranial Dissection The intradural stage of the transcondylar and supracondylar extensions of the far lateral approach was used for the
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2016.04.023
ORIGINAL ARTICLE ARNAU BENET ET AL.
photographic report (Figure 2). In the 2-D photograph (Figure 2 [left panel]), the vertebral artery (VA) appeared to be resting on the lateral mass of the C1. However, in the 3-D photograph (Figure 2 [right panel]), where the perspective of depth is present, the relative distance between the 2 structures was distinctly emphasized, showing the VA free from C1 and retracted off the surgical field. Moreover, the spatial orientation of the different cranial nerves (CN VIIeXII) in relation to each other and to the cerebellar hemisphere was better appreciated in the 3-D photograph. The depth from the condyle to CN XII, one of the most important concepts in efficient drilling during the transcondylar far lateral approach, was optimally appreciated in the 3-D photograph. The spatial relationship between CN XI and XII,
3-D IMAGING IN NEUROSURGERY
critical for drilling the jugular tubercle safely, was best understood in the 3-D photograph. Endoscopic Endonasal 3-D Photography The ventral compartment of the posterior fossa was exposed using an endoscopic endonasal transclival approach (Figure 3A and B). The 3-D photographs of the endoscopic endonasal transclival approach (Figure 3 [right panels]) provided a clear sense of depth of the surgical corridor compared with the 2-D photographs of the same exposure (Figure 3 [left panels]). The relative position of the pterygopalatine ganglion, the vidian nerve, and the internal carotid artery, critical for safety of the approach, could be understood efficiently in the 3-D
Figure 3. (A and B) Endoscopic endonasal dissection using the transclival approach. The 2-dimensional images are depicted on the left with their 3-dimensional (3-D) counterparts on the right. The different levels and depths at which the branches of the basilar artery originate (e.g., superior cerebellar artery, posterior cerebral artery, and anterior inferior cerebellar artery) were clearly appreciated in the 3-D image compared with the 2-dimensional image. The 3-D image highlighted the spatial relationship between the abducens nerve (cranial nerve VI) and the anterior inferior cerebellar artery, showing that cranial nerve VI traversed under the anterior inferior cerebellar artery near its origin at the pontomedullary sulcus. The relative position between the pterygopalatine ganglion and the internal carotid artery was also better understood in the 3-D image. AICA, anterior inferior cerebellar artery; BA, basilar artery; ICA, internal carotid artery (L, lacerum segment; PC, paraclival segment); IMA, internal maxillary artery; IOn, infraorbital nerve; PCA, posterior cerebral artery; Pg, pituitary gland; SCA, superior cerebellar artery; ON, optic nerve; VA, vertebral artery.
WORLD NEUROSURGERY 91: 317-325, JULY 2016
www.WORLDNEUROSURGERY.org
321
ORIGINAL ARTICLE ARNAU BENET ET AL.
322
www.SCIENCEDIRECT.com
3-D IMAGING IN NEUROSURGERY
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2016.04.023
ORIGINAL ARTICLE ARNAU BENET ET AL.
photograph (Figure 3B). Also, the depth and size of the paraclival segment of the internal carotid artery (Figure 3A) and its relationship in depth to CN VI (Figure 3B), which are very relevant to this step of the approach, were best shown in the 3-D photograph. The 3-D photograph in Figure 3A (right panel) provided a better understanding of the difference in depth between the superior cerebellar artery and the pituitary gland. It also clearly depicted how the CN VI traversed in relation to the anterior inferior cerebellar artery near its origin at the pontomedullary sulcus, a relationship not clearly appreciated in the 2-D photograph. Additionally, the relative distance between the anterior bend of the internal carotid artery and the posterior cerebral artery and CN III was best illustrated in the 3-D photograph (Figure 3B [right panel]). Microsurgical Case Illustration A 41-year-old woman presented with a recent history of imbalance on her right side, with multiple falls and garbled speech (Figure 4AeF). Angiography (Figure 4F [left panel]) revealed complete occlusion of both the proximal basilar artery and the left VA with a small right VA and absence of the left posterior communicating artery. An extended retrosigmoid approach for a V3 VA-to-anterior inferior cerebellar artery bypass with a saphenous vein graft was selected for blood flow augmentation in the posterior circulation (Figure 4AeE). Postoperative angiography (Figure 4F [right panel]) showed a patent bypass with robust restoration of blood flow. The patient had an uneventful postoperative course and remained symptom-free at postoperative 6-month follow-up. Figure 4AeE depicts 2-D intraoperative photographs in the left panel and the corresponding 3-D intraoperative photographs in the right panel. As evidenced in Figure 4, intraoperative 3-D photographs provided the opportunity for understanding the depth of the surgical corridor and allowed for the identification of relative distance and depth between CN XI and the anterior inferior cerebellar artery and the real angulation and position of the microinstruments, all of which were suboptimally shown through traditional 2-D imaging. The 3-D images allowed better appreciation of the positioning of the needle and the placement of suture throws on either side of the lumen of the interposition graft (Figure 4B and C), which was not adequately conveyed in the corresponding 2-D images, where the vessels appear flat. The 3-D images also provided better comprehension of the angulation and positioning of the clip relative to the surrounding structures. However, in terms of picture resolution, high-resolution 2-D and
3-D IMAGING IN NEUROSURGERY
3-D still photography using the macroscopic method, similar to that reported in Figure 2, were superior to the 3-D photographs obtained using the screenshot function during intraoperative microsurgical recording (Figure 4).
DISCUSSION Binocular vision is based on the principle that both eyes, situated at a small distance apart, view a target in a slightly different manner from each other. The 2 images, corresponding to each eye, are referred to as the “stereoscopic pair” and are superimposed to form a 3-D image, enabling depth perception. Depth perception is key in interpreting and understanding life for most of our daily tasks, but it becomes critical during surgery, as various structures are located in close proximity to each other, requiring skillful maneuvering of instruments to avoid unintentional damage to important neurovascular structures. This concept is best summarized by 2 axioms: “the better we see, the more we know,” and “the more we know, the more we see.”22 It is essential for the neurosurgical trainee to develop an excellent visual and tactile understanding of the surgical field. The trainee must learn to deftly navigate the surgical window to minimize complications and to ensure the patient’s optimal recovery and survival. Our side-by-side comparison of 2-D and 3-D recording (Figures 2e4) highlights the importance of 3-D imaging in providing depth perception and discerning closely related structures. Depth perception is important in all surgical procedures, but it takes on added importance when endoscopic endonasal techniques are undertaken, as demonstrated in our example using the transclival approach (Figure 3A and B). To our knowledge, this is the first report in the literature describing the method to obtain endoscopic endonasal 3-D images of neurosurgical procedures using currently available 2-D endoscopes. The surgical corridor of most endoscopic approaches is very deep and inaccessible to the naked eye, as it is situated beyond the nares. The relative sensation of depth of field, essential to endoscopic endonasal surgery, is felt indirectly by the surgeon, through dynamic movements of the endoscope and haptic feedback from accompanying instruments. This sensation of depth is unique to the surgeon and is very difficult to convey in 2D videos or photographs. Because 3-D vision bypasses the reliance on depth of field by touching, the trainee is no longer dependent on tactile sensation alone to learn, potentially shortening the learning curve, as validated previously in literature.8 Shah et al.7 reported that the use of the 3-D endoscope was preferred by novices compared with the 2-D endoscope and
Figure 4. (AeE) Intraoperative images for end-to-side anterior inferior cerebellar artery and V3 bypass using a saphenous vein graft. The 2-dimensional images are depicted on the left with their 3-D counterparts on the right. (F) Preoperative and postoperative angiography images are depicted in the left and right panels, respectively. Preoperative angiography revealed complete occlusion of the proximal basilar artery and the left vertebral artery with a small right vertebral artery. Postoperative angiography shows successful bypass with the restoration of blood flow; the star denotes the position of the graft. The 3-D image provided better appreciation of the relative distance and depth between cranial nerve XI and the anterior inferior cerebellar artery and the real angulation and position of the clip and microinstruments. In (B) and (C), compared with the 2-dimensional image, in which the vessel appeared flat, the 3-D image allowed better appreciation of the positioning of the needle and the placement of suture throws on either side of the lumen of the graft. AICA, anterior inferior cerebellar artery.
WORLD NEUROSURGERY 91: 317-325, JULY 2016
www.WORLDNEUROSURGERY.org
323
ORIGINAL ARTICLE ARNAU BENET ET AL.
3-D IMAGING IN NEUROSURGERY
helped improve depth perception, particularly during vascular dissection. Moreover, other studies have already revealed that 3-D neuroendoscopy is superior to 2-D endoscopy for the performance of various tasks, improving speed and efficiency and decreasing errors.6,8 However, despite the importance of 3-D recording for endoscopic endonasal procedures, use of 3-D endoscopic endonasal technology (Olympus, VisionSense, Endocam) is currently limited. This is due in part to decreased image resolution compared with 2-D endoscopic images, making it more difficult to discern important intraoperative structures and potentially leading to an increase in surgical complications. Furthermore, the sharp, quick, dynamic movements necessary for successful endonasal dissection may distort the 3-D experience, which works best with steady cameras or slow traveling; this might otherwise be described as intraoperative “motion sickness.” With ongoing advancements in endoscopic 3-D technology, we believe that these limitations can be overcome in the future, and the use of intraoperative 3-D endoscopic endonasal procedures is likely to become mainstream in teaching centers. Understanding the spatial relationships of structures is critical to deftly navigating the microsurgical corridor. Our case illustration using 3D intraoperative images for a VA-to-anterior inferior cerebellar artery bypass with a saphenous vein graft (Figure 4AeE) underscores the importance of 3-D imaging in understanding spatial relationships between different structures. The utility of 3-D intraoperative imaging in providing anatomic orientation, particularly to inexperienced trainees, has also been proven in previous reports. Rubino et al.2 studied the use of a 3-D intraoperative microscope during anterior circulation aneurysm surgery and compared it with traditional 2-D imaging among 30 physicians. They demonstrated that 3-D imaging was superior to the traditional 2-D still photographs in providing depth perception and improving the understanding of spatial relationships. Without depth perception provided by 3-D imaging, not only is the surgical trainee’s ability to process the visual field compromised, but also it becomes increasingly difficult for the instructor to properly emphasize the relative orientation of necessary surgical landmarks and key structures. Although cadaveric dissection is invaluable in gaining hands-on exposure to neurosurgical techniques, it does not prepare an inexperienced neurosurgical trainee to anticipate situations unique to surgery (e.g., bleeding that obscures the surgical field, brain expansion and volume changes, systolic-diastolic movements of the parenchyma).2 Observation of routine procedures using 3-D video enables trainees to observe the procedure from the eyes of the neurosurgeon, allowing for integration of maneuvers and the surgical technique from the unique perspective and view of the lead surgeon. Another advantage of 3-D video recording is that rare cases can be documented and used for research and teaching.23 Furthermore, in resource-poor areas where access to neurosurgical educators is limited, recorded videos can be used to introduce surgical trainees to techniques they would not otherwise be exposed to. With increasing opportunities to access and upload large amounts of data to cloud-based servers, the surgical classroom of the future may consist of a library of nearly every surgical approach recorded with 3-D imaging to be accessed by students worldwide. However, although 3-D
324
www.SCIENCEDIRECT.com
imaging helps increase familiarity with the anatomy, it is not a replacement for direct surgical exposure or hands-on surgical training. An important point underscored by our results is that in terms of picture resolution, 2-D still photography is superior to screenshots obtained from the built-in microsurgical 3-D video recording. This resolution gap may be explained by the size and density of the camera sensors and may be overcome as technology advances. However, there is no substitute for intraoperative 3-D recording, as the video playback allows the surgeon and trainee to move through the procedure in context, identifying key structures and their relationships as they are encountered during the procedure. Some surgeons may argue that the routine use of intraoperative 3-D microscopic documentation is costly and extravagant.24 We believe that incorporating a 3-D recording system in the OR is a wise investment, as it requires a one-time investment, with a large payback in educational opportunities, academic growth, and technical advancement. Teaching neurosurgical anatomy with 3-D imaging helps shorten learning curves, reducing the cost of training and improving surgical skill, which then improves patient outcomes and decreases overall health care costs. In their report, Barone et al.24 expressed the concern that the size of 3-D equipment may interfere with the routine functioning of the operating procedure. With the development of new technologies and incorporation of 3-D hardware and software in the intraoperative microscope, this limitation has been partly overcome. Our photograph of the OR setup at our institution and the floor-plan depict that with the use of state-of-the-art equipment with built-in 3-D cameras and wall mounted screens, the use of 3-D devices does not interfere with the activity in the OR. The wall-mounted 3-D display enables nurses and the surgical assistant to have a greater understanding of the surgical procedure and to anticipate surgical events and needs. Also, residents and students are able to observe the procedure from different standpoints in the OR without being in the way of the surgeon and nursing and anesthesia teams, as illustrated in Figure 1. In addition to the aforementioned practical applications of 3-D imaging, an emerging concept is computerized 3-D virtual reality simulation systems that render data from magnetic resonance imaging and computed tomography into virtual 3-D models. These systems are interactive and allow preoperative surgical simulation.25 Studies have shown that 3-D virtual reality systems are superior to traditional teaching methods (e.g., 2-D PowerPoint presentations) in teaching neuroanatomy to medical students.26 These systems have also been shown to be of value in training residents, allowing them to choose the appropriate approach specifically tailored to each patient.27 We suggest that 3-D images of anatomic dissections obtained using the methods we describe be used as an adjunct to enhance the virtual simulation experience, as these images provide a realistic depiction of the neurosurgical anatomy (e.g., brain dissection). CONCLUSIONS Use of 3-D imaging is invaluable in surgical anatomy research and education, as it enables novices and trainees to become oriented to neurosurgical anatomy and understand the key spatial
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2016.04.023
ORIGINAL ARTICLE ARNAU BENET ET AL.
3-D IMAGING IN NEUROSURGERY
relationships. 3-D recording should be the means by which the virtual classroom of surgical anatomy and procedures of the future is built, ensuring that medical students and surgical residents have access to the best and most relevant surgical education offered. Future advancements in this arena should focus on development of techniques for glasses-free viewing and improving the available 3-D virtual reality systems to set the stage
REFERENCES 1. Van Beurden MHPH, Van Hoey G, Hatzakis H, IJsselsteijn WA. Stereoscopic displays in medical domains: a review of perception and performance effects. In: Rogowitz BE, Pappas TN, eds. Proceedings of SPIE: Human Vision and Electronic Imaging XIV. Bellingham, WA: SPIE; 2009. 2. Rubino PA, Bottan JS, Houssay A, Salas López E, Bustamante J, Chiarullo M, et al. Three-dimensional imaging as a teaching method in anterior circulation aneurysm surgery. World Neurosurg. 2014;82:e467-e474. 3. Nakabayashi H, Shimizu K. Stereoscopic virtual realistic surgical simulation in intracranial aneurysms. Neurol India. 2012;60:191-197. 4. Christopher LA, William A, Cohen-Gadol AA. Future directions in 3-dimensional imaging and neurosurgery: stereoscopy and autostereoscopy. Neurosurgery. 2013;72(suppl 1):131-138. 5. Martins C, Ribas EC, Rhoton AL Jr, Ribas GC. Three-dimensional digital projection in neurosurgical education: technical note. J Neurosurg. 2015;123:1077-1080. 6. Fraser JF, Allen B, Anand VK, Schwartz TH. Three-dimensional neurostereoendoscopy: subjective and objective comparison to 2D. Minim Invasive Neurosurg. 2009;52:25-31. 7. Shah R, Leight W, Patel M, Surowitz JB, Wong YT, Wheless SA, et al. A controlled laboratory and clinical evaluation of a three-dimensional endoscope for endonasal sinus and skull base surgery. Am J Rhinol Allergy. 2011;25:141-144.
for the widespread use of 3-D imaging in surgical anatomy research and didactics. ACKNOWLEDGMENTS We express our gratitude to the body donors and their families, who through their altruism contributed to making this project possible.
ex-vivo laparoscopic tasks. J Endourol. 2012;26: 1085-1088. 11. Shimizu S, Tanaka R, Rhoton AL Jr, Fukushima Y, Osawa S, Kawashima M, et al. Anatomic dissection and classic three-dimensional documentation: a unit of education for neurosurgical anatomy revisited. Neurosurgery. 2006;58:E1000 [discussion: E1000]. 12. UCSF Skull Base & Cerebrovascular Laboratory. Available at: http://skullbaselab.ucsf.edu/. Accessed April 12, 2016.
approach to the clival and petroclival regions. Neurosurgery. 2009;65(6 suppl):42-50 [discussion: 50-52]. 21. Benet A, Prevedello DM, Carrau RL, RinconTorroella J, Fernandez-Miranda JC, Prats-Galino A, et al. Comparative analysis of the transcranial “far lateral” and endoscopic endonasal “far medial” approaches: surgical anatomy and clinical illustration. World Neurosurg. 2014;81:385-396. 22. Yas¸argil MG. Legacy of microneurosurgery: memoirs, lessons, and axioms. Neurosurgery. 1999; 45:1025-1092.
13. Tayebi Meybodi A, Benet A, Lawton MT. Microsurgical clipping of bilateral superior hypophyseal artery aneurysms through unilateral pterional craniotomy: 3-dimensional operative video [e-pub ahead of print]. Neurosurgery. http://dx.doi.org/10. 1227/NEU.0000000000001023.
23. Heath MD, Cohen-Gadol AA. Intraoperative stereoscopic 3D video imaging: pushing the boundaries of surgical visualisation and applications for neurosurgical education. Br J Neurosurg. 2012;26: 662-667.
14. Montemurro N, Benet A, Lawton MT. Clipping of ruptured residual anterior communicating artery aneurysm post endovascular coiling: 3dimensional operative video. Neurosurgery. 2016; 12:93-94.
24. Barone DG, Ban VS, Kirollos RW, Trivedi RA, Bulters DO, Ribas GC, et al. Alternative costeffective method to record 3D intra-operative images: a technical note. Br J Neurosurg. 2014;28: 819-820.
15. Benet A, Lawton MT. V3 vertebral artery to M2 middle cerebral artery bypass: 3-dimensional operative video [e-pub ahead of print]. Neurosurgery, 2015 Nov 4.
25. Bohm PE, Arnold PM. Simulation and resident education in spinal neurosurgery. Surg Neurol Int. 2015;6:33.
16. Benet A, Rincon-Torroella J, Lawton MT. Microsurgical clipping of a superior hypophyseal artery aneurysm: part 2: distal dural ring dissection and clipping: 3-dimensional operative video. Neurosurgery. 2015;11(suppl 2):359 [discussion: 359]. 17. Benet A, Rincon-Torroella J, Lawton MT. Microsurgical clipping of a superior hypophyseal artery aneurysm: part 1: intradural anterior clinoidectomy: 3-dimensional operative video. Neurosurgery. 2015;11(suppl 2):358 [discussion: 358].
8. Kawanishi Y, Fujimoto Y, Kumagai N, Takemura M, Nonaka M, Nakai E, et al. Evaluation of two- and three-dimensional visualization for endoscopic endonasal surgery using a novel stereoendoscopic system in a novice: a comparison on a dry laboratory model. Acta Neurochir (Wien). 2013;155:1621-1627.
18. Benet A, Lawton MT. Aneurysm transection and intra-aneurysmal clipping of a giant ophthalmic artery aneurysm: 3-dimensional operative video [e-pub ahead of print]. Neurosurgery, 2015 Jul 28.
9. Votanopoulos K, Brunicardi FC, Thornby J, Bellows CF. Impact of three-dimensional vision in laparoscopic training. World J Surg. 2008;32: 110-118.
19. Benet A, Rincon-Torroella J, Lawton MT, González Sánchez JJ. Novel embalming solution for neurosurgical simulation in cadavers. J Neurosurg. 2014;120:1229-1237.
10. Honeck P, Wendt-Nordahl G, Rassweiler J, Knoll T. Three-dimensional laparoscopic imaging improves surgical performance on standardized
20. de Notaris M, Cavallo LM, Prats-Galino A, Esposito I, Benet A, Poblete J, et al. Endoscopic endonasal transclival approach and retrosigmoid
26. Kockro RA, Amaxopoulou C, Killeen T, Wagner W, Reisch R, Schwandt E, et al. Stereoscopic neuroanatomy lectures using a threedimensional virtual reality environment. Ann Anat. 2015;201:91-98. 27. Agarwal N, Schmitt PJ, Sukul V, Prestigiacomo CJ. Surgical approaches to complex vascular lesions: the use of virtual reality and stereoscopic analysis as a tool for resident and student education. BMJ Case Rep. 2012;2012. Conflict of interest statement: The authors declare that the article content was composed in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Received 7 March 2016; accepted 6 April 2016
WORLD NEUROSURGERY 91: 317-325, JULY 2016
Citation: World Neurosurg. (2016) 91:317-325. http://dx.doi.org/10.1016/j.wneu.2016.04.023 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2016 Elsevier Inc. All rights reserved.
www.WORLDNEUROSURGERY.org
325