Three-dimensional reconstruction of the topographical cerebral surface anatomy for presurgical planning with free OsiriX Software.

TECHNIQUE ASSESSMENT

Three-Dimensional Reconstruction of the Topographical Cerebral Surface Anatomy for Presurgical Planning With Free OsiriX Software Mehmet V. Harput, MD Pablo Gonzalez-Lopez, MD, PhD ur Tu¨re, MD Ug Department of Neurosurgery, Yeditepe University School of Medicine, Istanbul, Turkey Correspondence: U gur Tu¨re, MD, Department of Neurosurgery, Yeditepe University Hospital, Devlet Yolu Ankara Cad. No: 102/104, 34752 Kozyatagi-Istanbul, Turkey. E-mail: [email protected] Received, November 22, 2013. Accepted, March 3, 2014. Published Online, March 21, 2014. Copyright © 2014 by the Congress of Neurological Surgeons.

BACKGROUND: During surgery for intrinsic brain lesions, it is important to distinguish the pathological gyrus from the surrounding normal sulci and gyri. This task is usually tedious because of the pia-arachnoid membranes with their arterial and venous complexes that obscure the underlying anatomy. Moreover, most tumors grow in the white matter without initially distorting the cortical anatomy, making their direct visualization more difficult. OBJECTIVE: To create and evaluate a simple and free surgical planning tool to simulate the anatomy of the surgical field with and without vessels. METHODS: We used free computer software (OsiriX Medical Imaging Software) that allowed us to create 3-dimensional reconstructions of the cerebral surface with and without cortical vessels. These reconstructions made use of magnetic resonance images from 51 patients with neocortical supratentorial lesions operated on over a period of 21 months (June 2011 to February 2013). The 3-dimensional (3-D) anatomic images were compared with the true surgical view to evaluate their accuracy. In all patients, the landmarks determined by 3-D reconstruction were cross-checked during surgery with high-resolution ultrasonography; in select cases, they were also checked with indocyanine green videoangiography. RESULTS: The reconstructed neurovascular structures were confirmed intraoperatively in all patients. We found this technique to be extremely useful in achieving pure lesionectomy, as it defines tumor’s borders precisely. CONCLUSION: A 3-D reconstruction of the cortical surface can be easily created with free OsiriX software. This technique helps the surgeon perfect the mentally created 3-D picture of the tumor location to carry out cleaner, safer surgeries. KEY WORDS: 3-D reconstruction, Cortical venous anatomy, Gyri, Intrinsic brain tumor, OsiriX, Sulci Operative Neurosurgery 10:426–435, 2014

G

WHAT IS THIS BOX? A QR Code is a matrix barcode readable by QR scanners, mobile phones with cameras, and smartphones. The QR Code above links to Supplemental Digital Content from this article.

liomas develop in the white matter and initially grow by expanding into the gyral subcortical regions.1 Generally, these tumors tend to grow centrally toward the ventricular system, and it is in the advanced stages when the tumor invades different functional compartments and adjacent gyri.1,2 The surgeon’s strategy is usually to directly attack the tumor from its most approachable area to create ABBREVIATION: ICG, indocyanine green Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Website (www.neurosurgery-online.com).

426 | VOLUME 10 | NUMBER 3 | SEPTEMBER 2014

DOI: 10.1227/NEU.0000000000000355

a working space to remove the lesion. Therefore, neurosurgeons should have a complete understanding of the local morphological anatomy and the functional organization of the cerebral architecture. Many craniometric references have been correlated with different underlying gyri and sulci, and accurate knowledge of these references is helpful in assessing the extracranial projection of any given tumor.3,4 Once the craniotomy is done and the dura opened, it is important to identify the pathological gyral tissue as well as the sulci and gyri surrounding the lesion. Nevertheless, in many cases, these lesions cannot be visualized once the dura is opened because of the arachnoid thickness, an apparently normal cortical surface, or both. Image-guided surgery and the first

www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

OSIRIX AS PREOPERATIVE PLANNING TOOL

neuronavigation systems were introduced to resolve these limitations, among others, but these devices still have serious practical limitations for intrinsic tumors because of the problem of brain shift.5-8 Few commercially available software packages can be used to plan surgery because they lack the capability to automate a 3-dimensional (3-D) reconstruction of the cerebral surface from magnetic resonance images.9-14 In addition, such software is very expensive, and some packages do not allow the option of producing a 3-D image of the cerebral surface that includes vessels. Furthermore, most of these packages were developed for cranial base surgeries, which are forgiving of imperfect images. In our surgical approach to subcortical lesions, we recently incorporated an original, free, and user-friendly method to create 3-D reconstructions of our patients’ brains, allowing us to preoperatively study the cerebral surface and exact topography of any given lesion, as well as the relationships with the surrounding vascular and sulcal landmarks.

METHODS Patients After obtaining institutional review board approval, from June 2011 to February 2013, a series of 51 consecutive patients with the diagnosis of an intrinsic supratentorial neocortical brain lesion were studied preoperatively and included in this image-processing protocol.

Image Acquisition and Processing All patients underwent scans with a 3-T magnetic resonance scanner (Achieva; Philips, Eindhoven, the Netherlands) with an 8-channel head coil. The sequences required for high-resolution 3-D reconstructions were 3-D T1 turbo fast echo, a 1-mm slice thickness, and images both nonenhanced and enhanced with gadolinium, allowing the creation of accurate multiplanar reconstructions. All image data were entered into the local net server in the Digital Imaging and Communications in Medicine (DICOM) file format. The data were then imported and processed through free medical imaging software (OsiriX Medical Image software, version 3.3.2; www.osirix-viewer.com)15 running on a MacBook Pro 2.53 GHz Intel Core 2 Duo (Apple Computer, Inc, Cupertino, California).

3-D Reconstruction of the Cerebral Surface To describe the process of 3-D reconstruction, step-by-step instructions are described here for 2 illustrative cases.

Patient 1 A 44-year-old woman presented at admission with tonic-clonic generalized seizures. Magnetic resonance imaging (MRI) showed a nonenhancing lesion located in the posterior part of the left superior frontal gyrus (F1), pushing posteriorly the superior aspect of the precentral gyrus (Figures 1A-1C). Once the DICOM files for the patient were loaded into the OsiriX database, the 3-D T1-weighted sequences with and without gadolinium were processed. Sequences without gadolinium were used to delineate the cerebral surface with high accuracy, reproducing the sulcal

FIGURE 1. Preoperative 3-dimensional T1 turbo fast echo axial (A), coronal (B), and sagittal (C) magnetic resonance images of patient 1 showing a lesion located in the posterior part of the left superior frontal gyrus pushing posteriorly the superior aspect of the precentral gyrus.

OPERATIVE NEUROSURGERY

VOLUME 10 | NUMBER 3 | SEPTEMBER 2014 | 427


HARPUT ET AL

FIGURE 2. A, screenshot from the OsiriX software showing where the 3-dimensional (3-D) volume-rendering tool is. B, second screenshot from the OsiriX software. When the 3-D volume-rendering button is clicked, a 3-D image of the patient appears on the screen. C, third screenshot from the OsiriX software. After cropping half of the image, the border between the brain surface and cranium becomes clear. If the image does not alternate, the window level helps produce a clear image. The area covered with a green line represents the area to be removed with a sculpting tool. D, the fourth screenshot from the OsiriX software demonstrates how the fly-through point helps to acquire images of the same position and size from enhancing and nonenhancing 3-D images, with and without vessels, respectively.

and gyral architecture in detail and thus demarcating the tumor and its relation with the gyral surface. In addition, gadolinium-enhanced sequences were used to appreciate the anatomic relationship between the cortical vessels and the underlying cerebral cortex. The 3-D reconstruction process started with double-clicking the desired sequence in the local database window so that it loaded into the 2-dimensional (2-D) viewer window. In this new screen, we selected 3-D Volume Reconstruction from the 2-D/3-D Reconstruction Tools menu (Figure 2A). If the program presented a prompt in the 3-D Presets menu, we chose high contrast. Then, a 3-D reconstruction of the whole cranial space appeared on the main screen (Figure 2B). Brightness and contrast were adjusted to achieve the desired color and tissue density. The next step consisted of peeling away the skin, fat, bone, and dura to show only the cortical neural and vascular tissues. During this step, cropping the contralateral side of the head helped to identify the layers of the scalp when viewed from medial to lateral planes. This process was always done with only a few tools, including the sculpt, move, and rotate tools (Figure 2C). The last step was to orient the volumetric 3-D reconstruction simulating the expected surgical position. This position of the 3-D figure


can be saved by adding it as the fly-through point and exporting and saving it in the desired folder. When the same procedure is done with the gadolinium-enhanced sequences, after the superficial layers of the scalp are peeled back, this file can be reimported as the fly-through point. This trick matches the new 3-D figure to the same position and size as the previous 3-D figure (Figure 2D). Moreover, when these 2 images are exported as .jpg files, the gadolinium-enhanced image can be placed over the nonenhanced image with any image-processing software, such as Photoshop or any presentation software such as Keynote or PowerPoint. Then, by decreasing and increasing the opacity of the upper image, the dense vascular cortical net becomes the surgical landmark for realizing the sulcal and gyral anatomy surrounding the tumor. In this patient, the 3-D reconstruction showed a significant enlargement of the posterior part of the F1 at the level of its connecting arm to the precentral gyrus, which was moved posteriorly and laterally (Figure 3A). The precentral sulcus, the superior frontal sulcus, and the posterior frontal vein draining into the superior sagittal sinus were identified (Figure 3B). Thus, the cortical projection of the tumor was theoretically delineated between the superior frontal sulcus and the interhemispheric fissure and anteriorly by an anterior branch of the posterior frontal vein;




FIGURE 3. A, 3-dimensional (3-D) reconstruction of patient 1 showing the sulcal and gyral architecture of the left superior frontal gyrus and adjacent paracentral lobule and its relationship to a lesion centered in the posterior aspect of the superior frontal gyrus (F1) and extending posteriorly to the precentral gyrus (prg) through the precentral sulcus (prs). B, 3-D reconstruction with vessel implementation showing the close relationship of the lesion, the posterior frontal vein (pfv), and the superior sagittal sinus (sss). C, intraoperative image of the same patient in which a clear match with the previous 3-D reconstructions was achieved. The precentral sulcus and superior frontal sulcus (f1), as well as the sulcal and gyral architecture of F1 and their relationship to the subcortical lesion, are now easily distinguished. D, intraoperative image showing the surgical field after removal of the tumor. The tumor was removed subpially through a small pial incision. E, intraoperative image showing the surgical field after closure of the pia mater with 10-0 sutures. cs, central sulcus; F2, middle frontal gyrus; ihf, interhemispheric fissure; pog, postcentral gyrus.

however, the tumor extended into the superior aspect of the precentral gyrus. A left frontoparietal parasagittal craniotomy was done with the patient in supine position with a 15° head elevation. After opening the dura, the posterior frontal vein and its anterior thin tributary vein were easily


identified, matching accurately with the previous 3-D reconstruction (Figure 3C). The sulcal and gyral anatomy was then better understood. By using direct electrical cortical stimulation, the cortical map in the surgeon’s mind was verified. Because the 3-D reconstruction in conjunction with data from the functional MRI showed exactly where



HARPUT ET AL

FIGURE 4. Postoperative 3-dimensional T1 turbo fast echo. Axial (A), coronal (B), and sagittal (C) magnetic resonance images showing gross total resection of the tumor.

to stimulate the tissue to find the motor area, this stimulation took much less time than usual. The next stage consisted of checking these surface landmarks in the depth with high-resolution ultrasound (Prosound Alpha 10; Aloka, Tokyo, Japan). This tool showed a clear, deep, posterior delimitation by the precentral sulcus and inferior delimitation by the superior frontal sulcus. The covering arachnoid membrane was longitudinally incised and opened, and a complete subpial excision was done (Figure 3D). Hemostasis was achieved, and a new ultrasound study was done to check the remaining gyral and sulcal pattern. Then the pia-arachnoid was closed with a 10-0 suture to preserve the surgical field in the original anatomic condition (Figure 3E). In the early postoperative period, the patient had slight hemiparesis of the right arm, which fully resolved after a week. Postoperative MRI showed gross total resection of the tumor (Figures 4A-4C). Histopathological analysis showed an oligodendroglioma (World Health Organization grade II). The patient was discharged 4 days after the surgical procedure, and no adjuvant therapy was recommended. The step-by-step process of reconstructing the cortical surface with and without the vessels is provided (see Video, Supplemental Digital Content, http://youtu.be/5cXpQMobFpc).

Patient 2 A 43-year-old woman experienced partial seizures. Her MRI showed a nonenhancing lesion located in the right pars opercularis of the inferior frontal gyrus (F3) (Figures 5A-5C). The patient was admitted to our institution for surgical removal of the lesion. The 3-D reconstruction showed an obvious enlargement of the right frontoparietal operculum of F3 mainly affecting the pars opercularis and the posterior half of the pars


triangularis. The ascending ramus of the sylvian fissure was easily recognized (Figure 6A). A frontoparietal tributary vein crossing just over the pars opercularis of F3 was identified joining the superficial sylvian vein some millimeters posterior to the sylvian point (Figure 6B). This venous pattern represented the main preoperative anatomic landmark. The precentral sulcus and gyrus and the inferior frontal sulcus were also identified delimiting the tumor extension posteriorly and superiorly, respectively. The operation was done with the patient in the supine position with her head rotated 30° to the left. A right pterional craniotomy was carried out, and, after the dura was opened, the frontoparietal, orbitofrontal, and a portion of the temporal operculum were exposed surrounding the sylvian fissure (Figure 6C). During the venous phase of indocyanine green (ICG) videoangiography, a large frontoparietal tributary vein and the distal origin of a frontal midline-bridging vein were demonstrated (Figure 6D). The previous 3-D images were easily and quickly matched to the exposed surgical field and ICG videoangiography, and the tumor location was defined between the inferior frontal sulcus and its extension into the triangular aspect of the frontoparietal operculum, sylvian fissure, and precentral sulcus. Next, high-resolution ultrasound was used to check these surface landmarks in the depth. The ultrasound images showed a clear, deep posterior delimitation by the precentral sulcus, superior delimitation by the inferior frontal sulcus, and inferior limiting by the sylvian fissure. The arachnoid was opened and a complete subpial resection was achieved. Once the tumor was removed, pertinent hemostasis was carried out, and the arachnoid was approximated with a 10-0 suture (Figure 7A). The surgical field was inspected again with ultrasound and ICG videoangiography to check the anatomic result (Figure 7B). ICG videoangiography was also




FIGURE 5. Three-dimensional T1 turbo fast echo. Axial (A), coronal (B), and sagittal (C) magnetic resonance images of patient 2. These images show a frontoparietal opercular glial tumor on the right side that does not enhance.

useful to help the surgeon maintain precision in his work, as it clearly shows whether the cortical areas have been bruised during surgery. In this patient, there was no such area. The patient had no neurological impairment, and postoperative MRI showed no residual tumor (Figures 8A-8C). Histological analysis showed an oligodendroglioma (World Health Organization grade II), and no adjuvant therapy was suggested.

RESULTS Preoperative and intraoperative findings matched with great accuracy in all 51 patients, reinforcing the mental concept the surgeon had and facilitating the pure tumor resections, sparing the surrounding healthy neurovascular structures, even for patients in whom there was no clear surface abnormality. Demographic data for the patients and the location of the lesions are given in detail in Table.

DISCUSSION When dealing with intrinsic cerebral lesions, it is important to mentally correlate the preoperative cortical morphology and topography with the intraoperative findings. Classically, this correlation is achieved with a combination of careful study of the 3-planar images that MRI offers and a strong knowledge of the sulcal and gyral cerebral patterns.16 Alternative methods to localize the cortical projection of these lesions include the use of neuronavigation systems. These devices allow intraoperative delineation of the tumor contours and surrounding structures


through a preoperative imaging dataset. Once the dura is opened and cerebrospinal fluid is drained, however, neuronavigation alone becomes less valuable in delineating the tumor’s borders in some cases. When the tumor is decompressed internally, neuronavigation is truly useless because of the problem of brain shift.5-7 A 3-D image showing the tumor location using landmarks that remain unchanged is most useful, especially for patients with multiple tumors. After the first tumor is removed, it is very difficult to locate the other tumors with only neuronavigation. With the method that we describe here, a precise 3-D reconstruction of the gyral, sulcal, and cortical vascular patterns can be easily created in a few minutes with free medical imaging software. This technique is not affected by brain shift, so this navigation concept is based on anatomic landmarks (cortical veins, arteries, and sulci). Despite cortical displacement, these landmarks move along with the navigation, maintaining their relative positions. During surgical planning to resect a deeply located cavernous malformation, Esposito et al17 presented a similar method, defining their transsulcal trajectory based on a 3-D reconstruction without the implementation of cortical vessels. However, the sulcal and gyral patterns are not always easily visualized once the dura is opened, and basing the surgical strategy only on the gyral surface morphology may lead the surgeon down erroneous corridors. In this sense, Nakajima et al18 reported the so-called vessel-to-vessel registration. These authors overlapped the videos of various surgical procedures with different preoperative 3-D



HARPUT ET AL

FIGURE 6. A, 3-dimensional (3-D) reconstruction showing the sulcal and gyral architecture of the right hemisphere operculi and its relationship to a lesion centered in the posterior aspect of the pars triangularis and pars opercularis. The precentral gyrus (prg) and precentral sulcus (prs), sylvian fissure (sf-a) and superficial sylvian vein, as well as the frontoparietal tributary, pars opercularis (op), pars triangularis (tr), pars orbitalis (or), and inferior frontal sulcus (f2) are now easily distinguished. B, 3-D reconstruction with vessel implementation showing the close relationship of the lesion and a tributary frontoparietal vein draining into the superficial sylvian vein. A frontal vein directed medially to drain into the sagittal sinus is also visualized. C, intraoperative image of the same patient showing a clear match with the previous 3-D reconstructions. D, screenshot of the surgical field during the indocyanine green videoangiography mainly showing the position of the prefrontal, precentral, and central arteries in the frontoparietal operculum and the posterior temporal artery in the temporal operculum, as well as a large frontoparietal tributary vein and the distal origin of a frontal midline bridging vein. The vascular anatomy exactly matches the preoperative 3-D reconstruction. F2, middle frontal gyrus; sp, sylvian point; cs, central sulcus; pog, postcentral gyrus; pos, post central sulcus; T1, superior temporal gyrus.

angiographic reconstruction screenshots, showing that the vessels visible on the surface of the brain provide a good map for navigation because they parallel the sulcal and gyral paths and are well delineated by phase-contrast magnetic resonance angiography. In our experience, the findings provided by ICG videoangiography reinforced our understanding of the vascular architecture, especially the venous patterns, which were used as the main anatomic landmarks. ICG videoangiography also forces the surgeon to maintain self-control during microdissection because it clearly shows the integrity of the veins. It also reveals any cortical contusion from unintended trauma. Combining these surface landmarks with the use of ultrasound allowed us to add a new dimension to our anatomic understanding of the tumor. The use of intraoperative ultrasound for intrinsic brain gliomas has been widely adopted and is now routinely used. Its main advantages are its invulnerability from brain shift and to provide real-time image guidance and ease of use.19,20 In our


experience, ultrasonography is the best navigation tool available, especially to delineate the tumor’s margins before attacking it. The biplanar images that these devices offer, however, require a 3-D understanding of the involved anatomic areas. The technique that we describe allowed us to achieve this 3-D understanding, making it easier to delineate the cortical topographic projection of any given intrinsic lesion. This strategy helped us confirm the most appropriate surgical approach with the ultimate goal of removing the lesion and minimizing any injury to the surrounding structures. This software also allows the user to preoperatively modify the orientation of the volumetric image to accurately simulate the expected surgical field. Thus, once the dura is opened, the surgeon can compare the intraoperative findings and, through a quick view, distinguish the gyral and sulcal pattern and its relationship to the underlying lesion, making tumor removal safer. The combination of ICG videoangiography and ultrasonographic findings, as well as data




FIGURE 7. A, intraoperative image showing the surgical field after the pure lesionectomy and closure of the pia using 10-0 sutures. B, the surgical field was inspected again using indocyanine green (ICG) videoangiography with the aim of checking the anatomic results of the surgery. ICG video angiography is also useful for controlling oneself, as it clearly reveals the cortical areas that have been bruised by traumatic surgery. In this case, there is no such area. cs, central sulcus; T1, superior temporal gyrus; or, pars orbitalis; F2, mid frontal gyrus; tr, pars triangularis; f2, inferior frontal sulcus; prs, precentral sulcus; prg, precentral gyrus.

from neuromonitoring in selected patients, allows the surgeon to mentally create a perfect 3-D picture of the tumor location and its margins, allowing cleaner, safer, and better surgeries. Another advantage of learning how to prepare 3-D images of the cortical surface with and without vessels appears when the surgeon prepares a presentation for an event. The images and videos produced with this software are most useful for clearly explaining

the anatomic specifications of any lesion and also, without a doubt, to impress the audience. We emphasize once more that this software is free. All a surgeon needs is a computer to run the software and 5 or 10 minutes to reconstruct the cortical topographical anatomy. Cooperative studies with computer engineers would definitely improve this method and can be done easily because the software is open source.

FIGURE 8. Postoperative 3-dimensional T1 turbo fast echo. Axial (A); coronal (B); sagittal (C) magnetic resonance images of patient 2 revealed no residual tumor.




HARPUT ET AL

TABLE. Demographic and Lesion Characteristics of Study Casesa Characteristics Sex Male Female Age, y Pediatric (average, 12.5 y) Adult (average, 49.2 y) Lesion side Left Right Lesion location Posterior F1 Middle F1 F2 Frontoparietal operculum T1 1 T2 1 T3 Superior parietal lobule 1 precuneus Inferior parietal lobule O1 1 O2 Lesion histopathology Cavernous malformation Dysembryoplastic neuroepithelial tumor (WHO grade I) Focal cortical dysplasia Ganglioglioma (WHO grade I) Pilocytic astrocytoma (WHO grade I) Diffuse astrocytoma (WHO grade II) Oligodendroglioma (WHO grade II) Oligoastrocytoma (WHO grade II) Anaplastic oligodendroglioma, anaplastic oligoastrocytoma, and anaplastic astrocytoma (WHO grade III) Glioblastoma and gliosarcoma (WHO grade IV) Metastasis

No.

%

30 21

58.8 41.2

6 45

11.8 88.2

30 21

58.8 41.2

10 3 4 8 17 4 3 2

19.6 5.9 7.8 15.7 33.3 7.8 5.9 4

4 2

7.8 4

1 1 1 6 8 1 4

2 2 2 11.8 15.7 2 7.8

18

35.1

5

9.8

CONCLUSION The freeware method presented in this technical note is a helpful intraoperative tool to delineate the cortical projection of neocortical tumors as well as other subcortical lesions that do not show any morphological change within the cortical architecture. This free, simple, and user-friendly software platform allows the surgeon to process all the image data in a few minutes through an easy toolbar function display. Its expansion in the neurosurgical community could help develop new uses, with the aim of improving our preoperative planning and surgical procedures and outcomes. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.

REFERENCES

a

F1, superior frontal gyrus; F2, middle frontal gyrus; T1, superior temporal gyrus; T2, middle temporal gyrus; T3, inferior temporal gyrus; O1, superior occipital gyrus; O2, middle occipital gyrus; WHO, World Health Organization.

Three-dimensional reconstruction of the cortical surface is easier and faster in the parasagittal and parietal areas, whereas the reconstruction takes more time in the temporal and pterional areas. This is due to the challenging bone removal part of the technique. Bone removal should be done without destroying the cortical surface. Therefore, when more irregularly shaped bone parts are involved, the bone removal part becomes more challenging, as in patient 2. The sculpt tool has 3 options triggered with 3 buttons. After selecting the region of interest, one may press the return button to delete the rest of the image while saving the region of interest or press the delete button to remove the selected region. The third and very useful button is the tab button, which puts back what was deleted to the selected region. So if the user removes more than needed or destroys the cortical surface, reselecting the removed area in different angles and pressing the tab key will do the trick.


Three-dimensional reconstruction of the cortical surface is not affected by the type or size of the subcortical lesion or cerebral edema. However, in meningioma cases, which are beyond the scope of this study, it is not always possible to reconstruct the cortical surface clearly. However, in these cases, detecting the extent of the tumor and the dural tail in 3-D is very useful in planning the size of the craniotomy. A 3-D reconstruction of the cortical surface is not clear enough in the reoperation cases due to meningocerebral adhesions. Three-dimensional reconstruction of the cerebral surface is a valuable addition to the neurosurgeon’s armamentarium, among other tools such as intraoperative ultrasonography, ICG videoangiography, neuronavigation, intraoperative MRI, and neuromonitoring.

1. Yas¸argil MG. Microneurosurgery, Volume IVA. Stuttgart: Thieme; 1994. 2. Yas¸argil MG. Microneurosurgery, Volume IVB. Stuttgart: Thieme; 1996. 3. Ribas GC, Yasuda A, Ribas EC, Nishikuni K, Rodrigues AJ Jr. Surgical anatomy of microneurosurgical sulcal key points. Neurosurgery. 2006;59(4 suppl 2):ONS177ONS210; discussion ONS210-ONS211. 4. Türe U, Yas¸argil DC, Al-Mefty O, Yas¸argil MG. Topographic anatomy of the insular region. J Neurosurg. 1999;90(4):720-733. 5. Roberts DW, Hartov A, Kennedy FE, Miga MI, Paulsen KD. Intraoperative brain shift and deformation: a quantitative analysis of cortical displacement in 28 cases. Neurosurgery. 1998;43(4):749-758; discussion 758-760. 6. Nabavi A, Black PM, Gering DT, et al. Serial intraoperative magnetic resonance imaging of brain shift. Neurosurgery. 2001;48(4):787-797; discussion 797-798. 7. Nimsky C, Ganslandt O, Cerny S, Hastreiter P, Greiner G, Fahlbusch R. Quantification of, visualization of, and compensation for brain shift using intraoperative magnetic resonance imaging. Neurosurgery. 2000;47(5):1070-1079; discussion 1079-1080. 8. Sun H, Roberts DW, Farid H, Wu Z, Hartov A, Paulsen KD. Cortical surface tracking using a stereoscopic operating microscope. Neurosurgery. 2005;56(suppl 1):86-97; discussion 96-97. 9. Mert A, Buehler K, Sutherland GR, et al. Brain tumor surgery with 3-dimensional surface navigation. Neurosurgery. 2012;71(2 suppl operative):ONS286-2ONS294; discussion ONS294-ONS295. 10. Beyer J, Hadwiger M, Wolfsberger S, Bühler K. High-quality multimodal volume rendering for preoperative planning of neurosurgical interventions. IEEE Trans Vis Comput Graph. 2007;13(6):1696-1703. 11. Chabrerie A, Ozlen F, Nakajima S, et al. Three-dimensional image reconstruction for low-grade glioma surgery. Neurosurg Focus. 1998;4(4):e7. 12. Ferroli P, Tringali G, Acerbi F, et al. Advanced 3-dimensional planning in neurosurgery. Neurosurgery. 2013;72(suppl 1):54-62.




13. Kockro RA, Serra L, Tseng-Tsai Y, et al. Planning and simulation of neurosurgery in a virtual reality environment. Neurosurgery. 2000;46(1):118135; discussion 135-137. 14. Oishi M, Fukuda M, Ishida G, Saito A, Hiraishi T, Fujii Y. Presurgical simulation with advanced 3-dimensional multifusion volumetric imaging in patients with skull base tumors. Neurosurgery. 2011;68(1 suppl operative):188-199; discussion 199. 15. Rosset A, Spadola L, Ratib O. OsiriX: an open-source software for navigating in multidimensional DICOM images. J Digit Imaging. 2004;17(3):205-216. 16. Ono M, Kubik S, Abernathey CD. Atlas of the Cerebral Sulci. Stuttgart: George Thieme; 1990. 17. Esposito V, Paolini S, Morace R. Resection of a left insular cavernoma aided by a simple navigational tool. Technical note. Neurosurg Focus. 2006;21(1):e16. 18. Nakajima S, Atsumi H, Kikinis R, et al. Use of cortical surface vessel registration for image-guided neurosurgery. Neurosurgery. 1997;40(6):1201-1208; discussion 1208-1210. 19. Gerganov VM, Samii A, Akbarian A, Stieglitz L, Samii M, Fahlbusch R. Reliability of intraoperative high-resolution 2-D ultrasound as an alternative to high-field strength MR imaging for tumor resection control: a prospective comparative study. J Neurosurg. 2009;111(3):512-519. 20. LeRoux PD, Berger MS, Ojemann GA, Wang K, Mack LA. Correlation of intraoperative ultrasound tumor volumes and margins with preoperative computerized tomography scans. An intraoperative method to enhance tumor resection. J Neurosurg. 1989;71(5 pt 1):691-698.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Website (www.neurosurgery-online.com).

Acknowledgment The authors thank Julie Yamamoto for editing the manuscript.

COMMENTS

T

he authors report their use of OsiriX, a free software package obtained from the Internet, to augment preoperative planning by reconstructing the brain surface in 3 dimensions. They describe this technique in a series of 51 cases consisting of a variety of brain pathologies in a variety of supratentorial locations. The time required to implement the technique is reported as 5 to 10 minutes; the potential steepness of the learning curve is not discussed, particularly for the subgroup of neurosurgeons who take slowly to computer-based technologies. Subcortical lesions are reported to have no impact on the imaging technology, but lesions involving the cortical surface, such as meningiomas, were not included, and meningocerebral adhesions negatively affect OsiriX application. The impact of subcortical lesions that present to the cortical surface was not discussed by the authors. I would be intrigued about the use of the technique in planning access to brainstem lesions, but infratentorial lesions are also not included in this study. Nevertheless, the OsiriX software appears to be a reasonable adjunct to consider for image-guided neurosurgery. N. Scott Litofsky Columbia, Missouri

T

he authors describe a relatively simple method for preoperative localization of the cortical anatomical landmarks. They analyzed the brain magnetic resonance imaging (MRI) of 51 patients with intrinsic supratentorial neocortical brain lesions using free OsiriX Medical Image software. The intraoperative cortical anatomical landmarks were visually matched with the OsiriX images and applied for navigation and localizing the lesion. Although the clinical outcome has not been depicted in this article, however, the authors believe that they had a 100% matching rate and have minimized the complications. They finally concluded that their


method is a helpful intraoperative tool to delineate the cortical projection of neocortical tumors, as well as other subcortical lesions without any cortical morphological changes. Brain tumor surgery with 3-dimentional (3-D) surface navigation using Stealth Viz neurosurgical planning software has been reported by Mert et al.1 They correlated superficial vessels and gyral anatomy on 3-D brain models with intraoperative images. There are also some other reports of 3-D visualization of surface anatomy using computed tomography (CT) scan, MRI and ultrasound (US).2-5 Rosset et al6 first introduced OsiriX as open-source software for navigating multidimensional DICOM images in 2004. More recently, Mandel et al reported their successful experiences with DICOM image processing using OsiriX as a navigation method in the Emergency cases.7 However, the current article is the first report of clinical application of this software for intra-axial neoplastic pathologies. The major advantage of this software is its widely and freely availability. It is not difficult to learn how to process images using this software and thus may be adoptable by clinicians without extensive computer knowledge. We welcome to all new innovative concepts that integrate the 3-D surface anatomical landmarks with neuronavigation, especially in situations that as a result of the brain shift, intra-operative neuronavigation becomes imprecise upon dural opening. However, we do not believe that using pre-operative reconstructed images derived from OsiriX will completely overcome the problem of brain shift, especially for localizing the subcortical lesions without cortical presentation. Navigation could be reliably updated with real time intraoperative images like Ultrasound without repeated patients registration.4-5 This updated navigation compensate for the effect of brain shift, allows better localization of the lesion and achieving maximum resection with lower complication rate.2,4,5 Finally, Three-dimensional reconstruction and visualization of the brain surface anatomical landmarks such as gyral and sulcal pattern as well as cortical vasculature could offer important navigational information. However, excellent surface accuracy is not a reliable indicator of the accuracy for deeper lesion. The authors still need to demonstrate the ability to account for potential dynamic changes upon dural opening. Moreover, the accuracy shown here has not been completely compared with the current intraoperative navigation methods. Parviz Dolati Alexandra Golby Boston, Massachusetts

1. Mert A, Buehler K, Sutherland GR, et al. Brain tumor surgery with 3-dimensional surface navigation. Neurosurgery. 2012;71(2 suppl operative):ons286-ons294. 2. Gerganov VM, Samii A, Akbarian A, et al. Reliability of intraoperative high-resolution 2D ultrasound as an alternative to high-field strength MR imaging for tumor resection control: a prospective comparative study. J Neurosurg. 2009;111(3):512-519. 3. LeRoux PD, Berger MS, Ojemann GA, Wang K, Mack LA. Correlation of intraoperative ultrasound tumor volumes and margins with preoperative computerized tomography scans. An intraoperative method to enhance tumor resection. J Neurosurg. 1989;71(5 pt 1):691-698. 4. Letteboer MM, Willems PW, Viergever MA, Niessen WJ. Brain shift estimation in image-guided neurosurgery using 3-D ultrasound. IEEE Trans Biomed Eng. 2005; 52(2):268-276. 5. Jodicke A, Springer T, Boker DK. Real-time integration of ultrasound into neuronavigation: technical accuracy using a light-emitting-diode-based navigation system. Acta Neurochir (Wien). 2004;146(11):1211-1220. 6. Rosset A, Spadola L, Ratib O. OsiriX: an open-source software for navigating in multidimensional DICOM images. J Digit Imaging. 2004;17(3):205-216. 7. Mandel M, Amorim R, Paiva W, et al. 3D preoperative planning in the ER with OsiriX®: when there is no time for neuronavigation. Sensors (Basel). 2013;13(5):6477-6491.



[Neurosurgical planning using osirix software].

Canine atlantoaxial optimal safe implantation corridors - description and validation of a novel 3D presurgical planning method using OsiriX™.

The anatomy of discharge planning.

Invasive electroencephalography monitoring: Indications and presurgical planning.

fetus during pregnancy by threedimensional ultrasonography using the HD live software: iconographic essay.

Anatomy of the middle temporal vein: implications for skull-base and craniofacial reconstruction using free flaps.

The NUKDOS software for treatment planning in molecular radiotherapy.

Topographical anatomy of the internal auditory canal. Implications for functional surgery in the cerebello-pontine angle.

Improved planning of endoscopic sinonasal surgery from 3-dimensional images with Osirix® and stereolithography.

Periorbital surface anatomy for the dermatologist.

Diffusion tensor tractography in the presurgical assessment of cerebral gliomas.

Surface anatomy of the ear.

Vascular anatomy of the gastric tube used for esophageal reconstruction.

Virtual reality AYRA software for preoperative planning in facial allotransplantation.

IML Software for Planning Internal Pilots.

Step-by-step surgical technique for mandibular reconstruction with fibular free flap: application of digital technology in virtual surgical planning.

Surface anatomy of the lip for the dermatologist.

Simultaneous Nanoscale Surface Charge and Topographical Mapping.

The topographical anatomy and blood supply of the carotid body region of the domestic fowl.

Repositioning template for mandibular reconstruction with fibular free flaps: an alternative technique to pre-plating and virtual surgical planning.

Variant adrenal vein anatomy: free with sample.

Microsurgical anatomy of the posterior cerebral artery.

Localization of specific language pathways using diffusion-weighted imaging tractography for presurgical planning of children with intractable epilepsy.

Blood oxygen level dependent functional magnetic resonance imaging for presurgical planning.