A novel cost-effective computer-assisted imaging technology for accurate placement of thoracic pedicle screws

Transcription

1 J Neurosurg Spine 15: , 2011 A novel cost-effective computer-assisted imaging technology for accurate placement of thoracic pedicle screws Technical note Yuichiro Abe, M.D., Ph.D., 1 Manabu Ito, M.D., Ph.D., 2 Kuniyoshi Abumi, M.D., Ph.D., 3 Yoshihisa Kotani, M.D., Ph.D., 1 Hideki Sudo, M.D., Ph.D., 2 and Akio Minami, M.D., Ph.D. 1 Departments of 1 Orthopaedic Surgery, 2 Advanced Medicine for Spine and Spinal Cord Disorders, and 3 Spinal Reconstruction, Hokkaido University Graduate School of Medicine, Sapporo, Japan Object. Use of computer-assisted spine surgery (CASS) technologies, such as navigation systems, to improve the accuracy of pedicle screw (PS) placement is increasingly popular. Despite of their benefits, previous CASS systems are too expensive to be ubiquitously employed, and more affordable and portable systems are desirable. The aim of this study was to introduce a novel and affordable computer-assisted technique that 3-dimensionally visualizes anatomical features of the pedicles and assists in PS insertion. The authors have termed this the 3D-visual guidance technique for inserting pedicle screws (3D-VG TIPS). Methods. The 3D-VG technique for placing PSs requires only a consumer-class computer with an inexpensive 3D DICOM viewer; other special equipment is unnecessary. Preoperative CT data of the spine were collected for each patient using the 3D-VG TIPS. In this technique, the anatomical axis of each pedicle can be analyzed by volumerendered 3D models, as with existing navigation systems, and both the ideal entry point and the trajectory of each PS can be visualized on the surface of 3D-rendered images. Intraoperative guidance slides are made from these images and displayed on a TV monitor in the operating room. The surgeon can insert PSs according to these guidance slides. The authors enrolled 30 patients with adolescent idiopathic scoliosis (AIS) who underwent posterior fusion with segmental screw fixation for validation of this technique. Results. The novel technique allowed surgeons, from office or home, to evaluate the precise anatomy of each pedicle and the risks of screw misplacement, and to perform 3D preoperative planning for screw placement on their own computer. Looking at both 3D guidance images on a TV monitor and the bony structures of the posterior elements in each patient in the operating theater, surgeons were able to determine the best entry point for each PS with ease and confidence. Using the current technique, the screw malposition rate was 4.5% in the thoracic region in corrective surgery for AIS. Conclusions. The authors found that 3D-VG TIPS worked on a consumer-class computer and easily visualized the ideal entry point and trajectory of PSs in any operating theater without costly special equipment. This new technique is suitable for preoperative planning and intraoperative guidance when performing reconstructive surgery with PSs. (DOI: / SPINE10721) Key Words pedicle screw placement accuracy of screw placement computer-assisted orthopedic surgery ipad scoliosis surgery thoracic spine Computer-assisted spine surgery is an innovative field that has rapidly developed in the past decade following the progress of information technology in general. Computed tomography navigation systems are one of the CASS units most commonly used to improve the accuracy of PS placement. 1,7,11,12,14,15,20,21,24 Detailed anal ysis of PS trajectories using a virtually developed 3D model and a CT navigation system enables surgeons to place PSs accurately irrespective of large anatomical Abbreviations used in this paper: AIS = adolescent idiopathic scoliosis; CASS = computer-assisted spine surgery; PS = pedicle screw; 3D-VG TIPS = 3D-visual guidance technique for inserting pedicle screws. variations. 7,11,14,15,20,21 Most navigation systems, however, require large special equipment and they lack portability. Therefore, computer-assisted surgery is performed at the limited number of facilities having such large and expensive hardware. The recent rapid development in information technology has enabled us to handle a larger amount of data than ever before, which has made precise 3D image analysis possible using only a typical consumer-grade computer. We developed a novel imaging technique to visualize the detailed anatomical configuration of a patient s This article contains some figures that are displayed in color on line but in black and white in the print edition. 479

2 Y. Abe et al. pedicle and to identify the optimal entry point and trajectory of the PS on a real 3D spine model. This new technique provides higher portability because it requires only commodity hardware and inexpensive software, such as a mobile laptop computer and a low-cost 3D DICOM viewer. We named this new imaging technique 3D-visual guidance technique for inserting pedicle screws (3D-VG TIPS). The purpose of this study was to introduce the concept of this novel imaging technique and to validate the technique in patients with AIS. Basic 3D-VG TIPS Essential Requirements of 3D-VG TIPS There are 3 essential requirements for this technique: 1) the patient s preoperative CT data set scanned with a thickness of < 1.0 mm; 2) hardware including a consumer-class laptop computer and monitor system; and 3) software including a 3D DICOM viewer and presentation software. The entire thoracic and lumbar levels (C7 S1) underwent CT scanning (Aquilion 64, Toshiba Medical Systems Corp.; SOMATOM Sensation 64, Siemens), and images were stored in DICOM format. The mean size of the data set was 524 ± 40 MB. We used a consumer-class laptop computer that had a Graphic Media Accelerator built-in Mobile Intel 945-GM express chipset, pendanttype displays, and ipad (Apple). A free 3D DICOM viewer (INTAGE Realia, KGT) and presentation software (PowerPoint, Microsoft) were used for analyzing PS entry points and trajectories and for generating intraoperative guidance slides. Analysis Method Using a 3D DICOM Viewer First Step: Determination of PS Direction. The ideal PS direction is determined by precise 3D analysis using a DICOM viewer. The basic manipulations described herein, such as 3D rotation and defining lines and planes, are generic operations common to most DICOM viewing software. In this technique, 3 coordinate axes are defined as follows: the x axis is horizontal and the y axis is vertical in the plane of the viewer window. The z axis is perpendicular to the x-y plane. Therefore, the cephalocaudal axis of the CT scan is not the same as the y axis in this method. Figure 1 illustrates a planning procedure for the left-side T-9 PS. The posterior elements just behind the pedicle are sliced, and coronal rotation of the vertebra is adjusted so that the upper endplate becomes parallel to the x axis (Fig. 1C). The spine is rotated 90 around the y axis, and the left lateral wall of the vertebra is seen (Fig. 1D). Clockwise rotation around the z axis is done so that the intended screw trajectory becomes parallel to the x axis (Fig. 1E), and then the spinal elements (above the blue dotted line in the image) are deleted. The vertebra is then rotated 90 around the y axis and x axis to obtain the accurate axial plane of the T-9 vertebra at the midpedicle level (Fig. 1G). An ideal screw trajectory in the axial plane (yellow dotted line) is adjusted to be parallel to the y axis by anticlockwise rotation around the z axis (Fig. 1H). The vertebra is rotated 90 around the x axis, and the cephalad elements of T-9 are visualized. This posterior image is now perpendicular to the intended axis of the PS (Fig. 1I). Second Step: Determination of the Entry Point on the Bony Surface. In the next step, the entry point of the PS on the bony surface is determined and the images for intraoperative guidance are created. Sequential crosssectional images of the pedicles from the bony surface to the vertebral body parallel to the previous image of Fig. 1I are produced by peeling off the most posterior surface (Fig. 2A); the obtained images are transferred to a presentation program (Fig. 2B). The center of the pedicle on the first slide, which is the deepest layer of the slides, is marked with a muzzle sight marker, and markers are pasted at the same point on every slide. The ideal entry point of the PS is automatically indicated on the bony surface by a muzzle sight marker in the final slide (Fig. 2B). Intraoperative Guidance Using the 3D-VG TIPS A TV monitor connected with a laptop computer is placed cephalad to the operating table and next to the fluoroscopy monitor (Fig. 3A), and the guidance slides obtained in the previous step are displayed on the TV monitor. A mobile monitor device such as an ipad can be used instead of a large TV monitor (Fig. 3B). An entry point is roughly identified by referring to anatomical bony landmarks such as the transverse process, superior facet joint, and lateral pars, and the precise location of the entry point is then determined in reference to the precise landmarks displayed on a TV monitor. Any calcified structures on the cortical surface, boundary line between cartilage of superior facet and lamina, mammillary processes, irregular bony prominence, and even Haversian canals can be used for landmarks (Fig. 4C and D). A bur hole is made in the cortical bone at the entry point using a 3.5-mm surgical bur. Pedicle probing is performed using the gearshift probing technique reported by Kim et al. 6 An oblique angle of the probing trajectory is also confirmed by checking the axial slices shown on the TV monitor. Recently, the ipad has been used to confirm the oblique trajectory of the PS by placing it next to the surgical site as shown in Fig. 3B. When an obvious perforation is detected by pedicle sounding after probing, hooks or sublaminar tapes are used instead of screw placement. 19,23 Illustrative Case This 14-year-old girl was diagnosed with idiopathic scoliosis; a single right convex thoracic curve was present. Her major curve was 63 from T-5 to L-1. Thoracic kyphosis, from T-5 to T-12, was 8. Her curve pattern was Lenke Type 1C. The PSs were placed from T-5 to L-1, except T-5 on the right side and T7 9 on the left side due to their narrowness demonstrated on preoperative CT scans (Fig. 4A and B). Planning of T-7 PS Placement on the Right Side Preoperative planning using the 3D-VG TIPS and an 480

3 Visual guidance technique for inserting pedicle screws Fig. 1. The concept of determining the direction of PS is shown. A: A coronal and sagittal rotation is roughly adjusted. B: The posterior column behind the pedicle is sliced and the upper endplate is indicated (red dotted line). C: The coronal rotation of the vertebra is adjusted. The upper endplate becomes parallel to the x axis. D: The vertebra is rotated 90 around the y axis to obtain the sagittal image. The intended screw trajectory is drawn as a blue dotted line. E: Clockwise rotation around the z axis is performed so that the blue dotted line becomes parallel to the x axis, and then the spinal elements above the blue dotted line are deleted. F: The vertebra is rotated 90 around the y axis. G: The vertebra is further rotated 90 around the x axis. Finally, an accurate axial plane of the T-9 vertebra is obtained. An ideal screw trajectory in the axial plane is indicated by the yellow dotted line. H: Anticlockwise rotation is done around the z axis. The yellow dotted line (the screw trajectory in the axial plane) must be completely parallel to the y axis. I: The vertebra is rotated 90 around the x axis and the cephalad elements are visualized. The T-9 vertebra on this image is perpendicular to the intended screw axis on the left side. operative photograph of T-7 on the right side are shown in Fig. 4C F. The pedicle was deformed into a V shape due to the deformity of the rib head. The entry point was carefully identified by referring to the transverse process, superior facet joint, ridgeline of the laminar cortex, and Haversian canal (Fig. 4C and D). Postoperative CT scans showed that the screw was properly placed at the bottom of the V-shaped pedicle without any misplacement. Validation of 3D-VG TIPS in Scoliosis Surgery Thirty consecutive patients with AIS who underwent posterior corrective surgery were evaluated. In 15 patients fluoroscopy-based PS insertion was performed without 3D-VG TIPS from 2008 to 2009 (fluoroscopy group). The remaining 15 patients received fluoroscopybased PS insertion combined with the 3D-VG TIPS from 2009 to 2010 (3D-VG group). There were no significant differences in the mean ages at surgery, the mean Cobb angles of major thoracic curve, and the mean numbers of instrumented spinal segments between the fluoroscopy group and 3D-VG group (Table 1). The mean operative times were 289 and 291 minutes in the fluoroscopy group and 3D-VG group, respectively. The mean preoperative planning time in the 3D-VG group was 37.4 minutes (5 minutes for creating a 3D model and 2 minutes to obtain images in the 3 coordinate axes for each pedicle). Screw misplacement was evaluated using postoperative axial CT scans. A screw with 2 mm of pedicle breach was considered malpositioned, in keeping with the description by Modi et al.10 The chi-square test was used for comparing the malposition rate between the groups. A p value < 0.05 was considered statistically significant. A total of 306 vertebrae (612 pedicles) were included in the fusion area, and 503 screws were inserted. The malposition rate in the thoracic spine was 11.6% in the fluoroscopy group and 4.5% in the 3D-VG group (p = 0.017). No medial perforation greater than 2 mm was observed in the 3D-VG group in either the thoracic or lumbar spine (Table 2). Discussion Several attempts have been made to improve the accuracy of thoracic PS placement in deformity correction surgery.4 7,10,14 18,20,21 In the present study, we introduced a new, affordable computer guidance technique that was designed to require only a common laptop computer and software. The modest computing performance requirements for 3DVG TIPS confer portability and allow surgeons to perform 481

4 Y. Abe et al. Fig. 2. Determination of the entry point on the posterior bony surface. A: From the posterior bony surface to the pedicle, any levels of the cross-sectional images can be obtained by peeling off the surface. B: The sequential images are embedded on each slide on the presentation program. The center of the pedicle on the first slide is marked with a muzzle sight marker, and markers are pasted in the same coordinate point on every slide. The entry point is indicated on the bony surface by a muzzle sight marker in the final slide. preoperative planning on a mobile computer anywhere, such as at the office, home, or even in transit. Surgeons can also bring the computer to the operating room and use the surgical plans as intraoperative guidance for placing PSs. The 3D-VG TIPS reduced the malposition rate of thoracic PSs in deformity correction of AIS from 11.6% to 4.5%. This result is superior to the malposition rates achieved with free-hand and fluoroscopy-guided methods (7% 25%)3,6,10,14,22 and compares favorably with those of CT-based navigation (1.8% 4.2%)7,11 or a fluoroscop ic navigation system (Iso-C) (0.8% 3%).14,21 A possible reason for the improvement of the malposition rate may Fig. 3. The surgical setting. A: Operative guidance slides are displayed on the TV monitor that is placed just cephalad to the operating table. Surgeons can easily identify the location of the entry point by looking at the 3D-VG TIPS monitor. Lateral fluoroscopy is used for checking the cephalocaudal location of the entry point and the sagittal direction of probing or tapping. B: A mobile monitor device (for example, an ipad) can be used instead of a TV monitor when there is no display system in the operating room. This kind of small monitor is helpful in determining the axial rotation angle of the screw. 482

5 Visual guidance technique for inserting pedicle screws Fig. 4. Studies obtained in a 14-year-old girl with AIS in whom the T5 L1 curvature was 63. A: Preoperative radiographs. B: Postoperative radiographs. Pedicle screws were inserted from T-5 to L-1 except in T-5 on the left side and T7 9 on the right. C: The entry point of the T-7 screw on the right side was indicated by a muzzle sight marker in the guidance slide. Smaller calcified structures are available as landmarks in the 3D-VG TIPS. In this case, the ideal entry point exists just cephalad to the Haversian canal (yellow arrow). D: Intraoperative photograph from the same angle. Medial-lateral orientation was easily identified by referring to the lateral edge of the superior facet joint (green arrow) and Haversian canal (yellow arrow). E: Merged image of panels C and D. Although the shape of the cortical bone was changed by the osteotomy, the bony surface around the entry point remained intact. F: Cross-sectional image at the pedicle level. We planned to place the screw at the bottom of the V-shaped pedicle. G: Postoperative CT scans. The screw was safely placed as planned preoperatively. be the new and accurate way of identifying the PS entry point. The real benefit of 3D-VG TIPS is the precise surface-rendering images that provide visualization of the PS entry point on the bone surface and the anatomical characteristics. Thus, the 3D-VG TIPS overcomes the drawbacks of conventional anatomical methods in defining the PS entry points by the common bony landmarks in each individual despite variations in individual and level-to-level anatomy.2,10,13 In addition to the identification of the ideal entry point for the screw, a correct insertion trajectory is also essential for accurate PS placement. There have been reports of several methods of trajectory guidance, such as computer-assisted navigation, using templates and angler jigs.1,8,9,14,21,24 Even without any specially designed tools other than the ipad for confirming axial trajectory of the PS during surgery, the screw malposition rates were significantly lower with the 3D-VG technique than with TABLE 1: Summary of demographic and operative data* PS Placement Group Variable Fluoroscopy 3D-VG Significance no. of patients (M/F ratio) mean age at op in yrs (range) mean Cobb angle of major thoracic curve (range) mean no. of instrumented segments (range) op time in mins (range) 15 (0:15) 13.9 (11 18) 60.5 ± 8.7 (46 72 ) 10.5 ± 2.6 (6 14) 289 ± 61 ( ) 15 (1:14) 15.4 (12 20) 60.6 ± 9.8 (47 83 ) 10.1 ± 1.8 (7 13) 291 ± 37 ( ) * Values are shown as the mean ± SD. Abbreviation: = not significant. 483

6 Y. Abe et al. TABLE 2: Summary of PS-related data Groups PS Placement Group Fluoroscopy 3D-VG p Value thoracic level screw perforation* medial lateral medial lateral Grade 0 (w/in pedicles) Grade 1 (0 2 mm) Grade 2 (2 4 mm) Grade 3 (4 6 mm) inserted screws (no.) malpositioned screws (no.) 23 9 malposition rate 11.6% 4.5% lumbar level screw perforation* medial lateral medial lateral Grade 0 (w/in pedicles) Grade 1 (0 2 mm) Grade 2 (2 4 mm) Grade 3 (4 6 mm) inserted screws (no.) malpositioned screws (no.) 2 0 malposition rate 3.4% 0% total (thoracic & lumbar) inserted screws (no.) malpositioned screws (no.) 25 9 malposition rate 10.6% 3.8% * Grades 0 and 1 = safely inserted screws; Grades 2 and 3 = malpositioned screws. p < the fluoroscopy technique. This fact suggests that if experienced surgeons can identify the ideal entry point of the PS, they can place the screws accurately by checking sequential cross-sectional images, also with the aid of an ipad, at any time during the procedure. In fact, in our early clinical series there was no medial breach exceeding 2 mm at any spinal level. In our early series, the 3D-VG TIPS did not increase the operation time because of its simple setting in the operative theater. The technique, however, required preoperative preparation to plan the ideal entry point and trajectory of each PS. In the present study, we confirmed that it took 40 minutes to assess 20 pedicles in each patient preoperatively, which was not long compared with CT navigation systems. The preparation time could be significantly shortened by decreasing the number of levels analyzed (for example, analyze only apical levels) or by producing automated new software customized for the 3D-VG TIPS. The O-arm and other intraoperative CT navigation systems do not require any preoperative preparation, and many studies have reported superior results with accurate PS placement. These high-tech systems, however, are still too costly so that not all spine surgeons can attain the benefits. Needless to say, precise preoperative assessment is the best way to ensure the safety of any kind of spinal surgery. By simulating screw insertion on virtual models of real patients on a laptop computer before surgery, a lead surgeon can obtain the best knowledge and feeling of how to properly place PSs, and this foreknowledge can be shared with other assisting surgeons. On our team, we make it a rule to discuss the appropriateness of each surgical plan for a patient before surgery and to review the surgical result of each case by checking the accuracy of PS placement on postoperative CT scans. This self-evaluation process could improve the quality of preoperative planning later on and remind each surgeon of his own tendencies for errors, both of which would reduce the risks of screw malposition in future surgeries. Conclusions The present study introduced a novel imaging technology (3D-VG TIPS) to visualize the ideal entry point and trajectory of PSs in the thoracic spine. Surgeons can perform precise preoperative planning of insertion 3-dimensionally on their personal mobile computers. This system can be used for preoperative planning anywhere, by surgeons with various levels of surgical experience, and can be brought into the operating room as a reliable intraoperative reference when placing PSs. Early clinical results have demonstrated that 3D-VG TIPS decreased PS malposition rates and prevented neurovascular complications in patients with scoliosis. Disclosure The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Author contributions to the study and manuscript preparation include the following. Conception and design: Abe. Acquisition of data: Ito, Abe. Analysis and interpretation of data: Abe. Drafting the article: Ito, Abe, Minami. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Ito. Statistical analysis: Abe. Administrative/technical/material support: Ito, Abumi, Kotani, Sudo. Study supervision: Ito, Abumi, Kotani, Sudo, Minami. Acknowledgments The authors thank Mr. Mason Mark for his support in Englishlanguage editing, and Drs. Yoshihiro Hojo and Shigeki Oshima for their support in data analysis. References 1. Anonymous: Full rotation three-dimensional intraoperative imaging during spinal surgery (O-arm Imaging System). OR Manager 24:after p 16, Broom MJ, Banta JV, Renshaw TS: Spinal fusion augmented by luque-rod segmental instrumentation for neuromuscular scoliosis. J Bone Joint Surg Am 71:32 44, Castro WH, Halm H, Jerosch J, Malms J, Steinbeck J, Blasius S: Accuracy of pedicle screw placement in lumbar vertebrae. Spine (Phila Pa 1976) 21: , Chung KJ, Suh SW, Desai S, Song HR: Ideal entry point for the thoracic pedicle screw during the free hand technique. Int Orthop 32: , Cinotti G, Gumina S, Ripani M, Postacchini F: Pedicle instrumentation in the thoracic spine. A morphometric and cadav- 484

7 Visual guidance technique for inserting pedicle screws eric study for placement of screws. Spine (Phila Pa 1976) 24: , Kim YJ, Lenke LG, Bridwell KH, Cho YS, Riew KD: Free hand pedicle screw placement in the thoracic spine: is it safe? Spine (Phila Pa 1976) 29: , Kotani Y, Abumi K, Ito M, Takahata M, Sudo H, Ohshima S, et al: Accuracy analysis of pedicle screw placement in posterior scoliosis surgery: comparison between conventional fluoroscopic and computer-assisted technique. Spine (Phila Pa 1976) 32: , Lu S, Xu YQ, Lu WW, Ni GX, Li YB, Shi JH, et al: A novel patient-specific navigational template for cervical pedicle screw placement. Spine (Phila Pa 1976) 34:E959 E966, Miyamoto H, Uno K: Cervical pedicle screw insertion using a computed tomography cutout technique. Technical note. J Neurosurg Spine 11: , Modi H, Suh SW, Song HR, Yang JH: Accuracy of thoracic pedicle screw placement in scoliosis using the ideal pedicle entry point during the freehand technique. Int Orthop 33: , Nakanishi K, Tanaka M, Misawa H, Sugimoto Y, Takigawa T, Ozaki T: Usefulness of a navigation system in surgery for scoliosis: segmental pedicle screw fixation in the treatment. Arch Orthop Trauma Surg 129: , Nottmeier EW, Young PM: Image-guided placement of occipitocervical instrumentation using a reference arc attached to the headholder. Neurosurgery 66 (3 Suppl Operative): , O Brien MF, Lenke LG, Mardjetko S, Lowe TG, Kong Y, Eck K, et al: Pedicle morphology in thoracic adolescent idiopathic scoliosis: is pedicle fixation an anatomically viable technique? Spine (Phila Pa 1976) 25: , Rajasekaran S, Vidyadhara S, Ramesh P, Shetty AP: Randomized clinical study to compare the accuracy of navigated and non-navigated thoracic pedicle screws in deformity correction surgeries. Spine (Phila Pa 1976) 32:E56 E64, Sakai Y, Matsuyama Y, Nakamura H, Katayama Y, Imagama S, Ito Z, et al: Segmental pedicle screwing for idiopathic scoliosis using computer-assisted surgery. J Spinal Disord Tech 21: , Suk SI, Lee CK, Kim WJ, Chung YJ, Park YB: Segmental pedicle screw fixation in the treatment of thoracic idiopathic scoliosis. Spine (Phila Pa 1976) 20: , Suk SI, Lee SM, Chung ER, Kim JH, Kim SS: Selective thoracic fusion with segmental pedicle screw fixation in the treatment of thoracic idiopathic scoliosis: more than 5-year followup. Spine (Phila Pa 1976) 30: , Takahashi J, Hirabayashi H, Hashidate H, Ogihara N, Kato H: Accuracy of multilevel registration in image-guided pedicle screw insertion for adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 35: , Takahata M, Ito M, Abumi K, Kotani Y, Sudo H, Ohshima S, et al: Comparison of novel ultra-high molecular weight polyethylene tape versus conventional metal wire for sublaminar segmental fixation in the treatment of adolescent idiopathic scoliosis. J Spinal Disord Tech 20: , Tanaka M, Nakanishi K, Sugimoto Y, Misawa H, Takigawa T, Nishida K, et al: Computer navigation-assisted spinal fusion with segmental pedicle screw instrumentation for scoliosis with Rett syndrome: a case report. Acta Med Okayama 63: , V Rajan V, Kamath V, Shetty AP, Rajasekaran S: Iso-C3D navigation assisted pedicle screw placement in deformities of the cervical and thoracic spine. Indian J Orthop 44: , Weinstein JN, Spratt KF, Spengler D, Brick C, Reid S: Spinal pedicle fixation: reliability and validity of roentgenogrambased assessment and surgical factors on successful screw placement. Spine (Phila Pa 1976) 13: , Winter RB, Denis F, Lonstein JE, Garamella J: Techniques of surgery, in Lonstein JE, Winter RB, Ogilvie JW, et al (eds): Moe s Textbook of Scoliosis and Other Spinal Deformities, ed 3. Philadelphia: WB Saunders, 1995, pp Zhang J, Weir V, Fajardo L, Lin J, Hsiung H, Ritenour ER: Dosimetric characterization of a cone-beam O-arm imaging system. J XRay Sci Technol 17: , 2009 Manuscript submitted October 18, Accepted June 7, Please include this information when citing this paper: published online July 15, 2011; DOI: / SPINE Address correspondence to: Manabu Ito, M.D., Ph.D., Department of Advanced Medicine for Spine and Spinal Cord Disorders, Hokkaido University Graduate School of Medicine, Kita-15-Jo, Nishi-7-Chome, Kita-ku, Sapporo , Japan. maito@ med.hokudai.ac.jp. 485