J Oral Maxillofac Surg 70:966-971, 2012
Splint Sterilization—A Potential Registration Hazard in Computer-Assisted Surgery Michael Figl, MS, DSc,* Christoph Weber, MS,† Ojan Assadian, MD, PhD,‡ Cyril D. Toma, MD, PhD,§ Hannes Traxler, MD,㛳 Rudolf Seemann, MD, DMD,¶ Godoberto Guevara-Rojas, MSc,# Wolfgang P. Pöschl, MD,** Rolf Ewers, MD, DMD, PhD,†† and Kurt Schicho, MD, DSc, PhD‡‡ Purpose: Registration of preoperative targeting information for the intraoperative situation is a crucial
step in computer-assisted surgical interventions. Point-to-point registration using acrylic splints is among the most frequently used procedures. There are, however, no generally accepted recommendations for sterilization of the splint. An appropriate method for the thermolabile splint would be hydrogen peroxide– based plasma sterilization. This study evaluated the potential deformation of the splint undergoing such sterilization. Deformation was quantified using image-processing methods applied to computed tomographic (CT) volumes before and after sterilization. Materials and Methods:
An acrylic navigation splint was used as the study object. Eight metallic markers placed in the splint were used for registration. Six steel spheres in the mouthpiece were used as targets. Two CT volumes of the splint were acquired before and after 5 sterilization cycles using a hydrogen peroxide sterilizer. Point-to-point registration was applied, and fiducial and target registration errors were computed. Surfaces were extracted from CT scans and Hausdorff distances were derived. Effectiveness of sterilization was determined using Geobacillus stearothermophilus. Results: Fiducial-based registration of CT scans before and after sterilization resulted in a mean fiducial registration error of 0.74 mm; the target registration error in the mouthpiece was 0.15 mm. The Hausdorff distance, describing the maximal deformation of the splint, was 2.51 mm. Ninety percent of point-surface distances were shorter than 0.61 mm, and 95% were shorter than 0.73 mm. No bacterial growth was found after the sterilization process. Conclusion: Hydrogen peroxide– based low-temperature plasma sterilization does not deform the splint, which is the base for correct computer-navigated surgery. © 2012 American Association of Oral and Maxillofacial Surgeons J Oral Maxillofac Surg 70:966-971, 2012
*Assistant Professor, Center for Biomedical Engineering and Physics, Medical University Vienna, Vienna, Austria. †PhD Student, Center for Biomedical Engineering and Physics, Medical University Vienna, Vienna, Austria. ‡Professor, Ernst-Moritz-Arndt-Universität Greifswald, Greifswald, Germany. §Associate Professor, Department of Orthopaedic Surgery, Medical University Vienna, Vienna, Austria. 㛳Assistant Professor, Center of Anatomy and Cell Biology, Medical University Vienna, Vienna, Austria. ¶Assistant Professor, Department of Craniomaxillofacial and Oral Surgery, Medical University Vienna, Vienna, Austria. #PhD Student, Department of Craniomaxillofacial and Oral Surgery, Medical University Vienna, Vienna, Austria.
**Assistant Professor, Department of Craniomaxillofacial and Oral Surgery, Medical University Vienna, Vienna, Austria. ††Professor, Chairman, Department of Craniomaxillofacial and Oral Surgery, Medical University Vienna, Vienna, Austria. ‡‡Associate Professor, Department of Craniomaxillofacial and Oral Surgery, Medical University Vienna, Vienna, Austria. Address correspondence and reprint requests to Dr Schicho: Department of Craniomaxillofacial and Oral Surgery, Medical University Vienna, Währinger Gürtel 18-20, 1090 Vienna, Austria; e-mail: [email protected]
© 2012 American Association of Oral and Maxillofacial Surgeons
FIGL ET AL
Computer-assisted surgical interventions have become indispensable within different medical fields. Registration of preoperative targeting information for the intraoperative situation is a crucial step. Among several procedures, point-to-point registration has proved to be most accurate.1 Technical realizations of this concept are fiducial markers attached to the patient, eg, microscrews inserted into the bone or reliable landmarks integrated into splints. Acrylic splints, supplied with fiducial markers, are often seen as advantageous compared with bone screws2 for clinical reasons (less invasive). Although similar splints have been used for many years,3-6 the influence of storage and sterilization on this basic registration tool has not been investigated. Maxillofacial surgical procedures are performed under aseptic conditions. Therefore, all surgical instruments and every medical device introduced into the surgical field must be sterile. For most instruments in trays used in computer-aided surgery, detailed reprocessing and sterilization methods are available (eg, ethylene oxide or steam sterilization). There are, however, no recommendations for sterilization of the thermolabile splint. Splint sterilization is therefore a potential hygienic deficit. Several sterilization procedures are in use, eg, heat (wet or dry), chemicals (ethylene oxide and those based on formaldehyde or glutaraldehyde), radiation (␤, ␥, or based on ultraviolet radiation), or laser-based apparatuses. Different sterilization processes have different characteristics and different effects on materials. All these methods have advantages and disadvantages, such as handling, reprocessing time, or the creation of possible toxic residuals. The release of toxic substances is also important for environmental protection. Using a low-temperature, low-pressure, vaporized aqueous hydrogen peroxide– based plasma sterilization method might circumvent this problem. This study evaluated deformation of the splint undergoing hydrogen peroxide– based plasma sterilization. Quantification was performed using image-processing methods applied to computed tomographic (CT) volumes before and after sterilization.
Materials and Methods To investigate deformation caused by hydrogen peroxide– based plasma sterilization, a CT scan of the splint was acquired before and after 5 low-temperature sterilization cycles. Quantitative evaluation was based on point-to-point registration and subsequent surface distance measurement of registered scans before and after sterilization. STUDY OBJECT
A navigation splint made of polymethyl methacrylate was used as the study object (Fig 1). Eight metal-
lic markers were placed in the side parts of the splint and used for registration. Because the splint was roughly symmetric, 4 small steel cylinders were placed on parts on the right side near the fiducial markers to avoid confusion on CT images. This kind of navigation splint is commonly used in computer-aided oral and maxillofacial surgery at the Medical University Vienna.7 A patient-specific occlusal stent is attached to the navigation splint. The patientspecific part, adapted to the teeth, is made of Folidur S (Styrol-Butadien-Styrol; RSB Duna Dental, Dunningen, Germany), and the connective part is manufactured from Paladur (Heraeus Kulzer, GmbH, Hanau, Germany). Six steel spheres were in the mouthpiece and used as targets for an approximation of the target registration error (TRE). STERRAD 100S STERILIZER
For this study, a Sterrad 100S sterilizer (Advanced Sterilization Products, Irvine, CA) was used. The basic principle of the Sterrad sterilizer is the germicidal property of hydrogen peroxide. Aqueous hydrogen peroxide is vaporized in a cylindrical chamber at low temperature and low pressure. The sterilizing action of aqueous hydrogen peroxide is caused by the lethal oxidizing capacity of viable elements on or in the micro-organism after condensation into the liquid phase.8,9 The Sterrad 100S with a 73-L sterilization chamber contains 2 shelves for placement of instruments to be sterilized within a cylindrical chamber. The sterilization cycle of the Sterrad 100S is 55 minutes. During the injection phase, a peak hydrogen peroxide concentration of 6 mg/mL is reached. The Sterrad 100S operates in 5 phases: a vacuum phase, an injection phase, a diffusion phase, a plasma phase, and a final ventilation phase. Vacuum Phase After the chamber door was closed and automatically sealed, the chamber was evacuated to 0.3-mm Hg pressure. Injection Phase A solution of hydrogen peroxide and water was injected into the evacuated sterilization chamber from a cassette inserted into the sterilizer, which evaporated the aqueous hydrogen peroxide solution and dispersed it into the chamber. The chamber temperature was controlled not to exceed 40°C to 45°C to decrease the chance of condensation. The chamber pressure increased slightly during the injection phase as the hydrogen peroxide evaporated. Relative humidity was 6% to 14%, and the equilibrium vapor pressure of water at 40°C was approximately 60 mm Hg.10 Diffusion Phase The hydrogen peroxide solution vaporized and diffused throughout the chamber, surrounding the
HYGIENE AND DEFORMATION OF NAVIGATION SPLINTS
FIGURE 1. A, Navigation splint with fiducial and target markers. The 8 fiducial markers are located at the side parts, and 6 targets can be found in the mouthpiece. B, Splint mounted on a patient during an intervention. Figl et al. Hygiene and Deformation of Navigation Splints. J Oral Maxillofac Surg 2012.
splint. During the diffusion phase (4 minutes in duration), the hydrogen peroxide vapor was allowed to permeate the chamber and completely expose all surfaces of the splint to the sterilant. At the completion of the diffusion phase, the chamber pressure was decreased to 0.5 mm Hg. Plasma Phase Radiofrequency energy was applied to the chamber to create an electromagnetic field, which caused the formation of cloudlike plasma of highly reactive molecules from the hydrogen peroxide vapor. The radiofrequency plasma discharge lasted for 15 minutes. In the plasma state, the hydrogen peroxide vapor broke up into reactive species that included free radicals. The combined use of hydrogen peroxide vapor and plasma safely and rapidly sterilized the splint without leaving
toxic residues. After the reaction, the activated components lost their high energy and recombined to form primarily oxygen, water, and other nontoxic byproducts.11 Ventilation Phase The radiofrequency energy was turned off, the vacuum was released, and air was filtered through a highefficiency particulate air filter entering the chamber to return it to normal atmospheric pressure. Then, the chamber was re-evacuated and vented again. The vapor purged from the chamber was vented to the atmosphere through a catalytic filter to decompose all remaining traces of hydrogen peroxide into water and oxygen vapor. The splint required no aeration and was immediately ready for use after the 55-minute cycle. The sterilization process was repeated 5 times in total, with
FIGL ET AL REGISTRATION
Point-to-point registration was applied using an implementation of the method by Horn12 using the GNU scientific library (Free Software Foundation, Inc, Boston, MA). Mean fiducial distance (fiducial registration error)13 and maximal distance were derived. TRE13 was computed using the 6 target markers at the mouthpiece. SURFACE GENERATION AND DISTANCE MEASUREMENT
FIGURE 2. The distance of a point x 僆 A to a path B is defined by d(x, B):⫽ infy 僆 B储x – y储. Figl et al. Hygiene and Deformation of Navigation Splints. J Oral Maxillofac Surg 2012.
Surfaces were generated from rescaled CT volumes using the 3D Slicer gray-scale model maker filter, an implementation of the marching cubes algorithm. To quantify correspondence between surfaces, the Hausdorff distance (dH) was used.14 It is defined using the maximal distance of a point in surface A to surface B (Fig 2). A symmetric definition is given by the equation: dH(A, B) : ⫽ max(supx 僆 A d[x, B], supx 僆 B d[y, A])
2 hours between each reprocessing cycle. The splint to be sterilized was used in a meticulously cleaned and dried condition, packed in cellulose- and latex-free Tyvek wrapping (Dupont, Wilmington, DE), and placed into metal grid baskets. Only materials and consumables recommended by the manufacturer were used. The sporicidal activity of every sterilization process was assessed using Geobacillus stearothermophilus ATCC 7953 (American Type Culture Collection, Manassas, VA) biological indicators. None of the indicators showed growth after the sterilization process. IMAGE ACQUISITION AND PREPROCESSING
Two CT volumes of the splint were acquired before and after 5 sterilization cycles using the Sterrad 100 S. The scans were obtained using a Siemens Sensation cardiac 64-slice CT scanner (Siemens, Erlangen, Germany) according to a standard dental scan protocol and a slice distance of 0.75 mm at the Prince Court Medical Centre (Kuala Lumpur, Federation of Malaysia). Before further use, the 2 datasets were resampled to generate cubic voxels with an edge length of 0.5 mm. Medical image processing (3D Slicer; Harvard Medical School, Boston, MA) was used for this purpose. LOCATION OF FIDUCIAL AND TARGET MARKERS
Center points of all fiducial and target markers were identified and their positions measured in the original and resized datasets. Scans were converted to Visualization Toolkit (Kitware Inc, Clifton Park, NY) volumes (http://www.vtk.org) using 3D Slicer. Thresholds for the volumes were then determined and connected components were labeled. The resulting components were manually classified into target and fiducial markers. Localization of the markers was determined by the center of mass of their respective components.
using d(x, B) : ⫽ infx 僆 B㛳x ⫺ y㛳 Because the Hausdorff distance decreases the distance information to the most exceptional point correlation, histograms of d(x, A), x 僆 B and d(x, B), x 僆 A are also provided. The Hausdorff distance and point distances were computed using the open-source software MESH15 applied to the registered surfaces before and after sterilization.
Results Calculating the point-based registration of fiducial markers on CT scans before and after sterilization resulted in a mean fiducial distance (fiducial registration error) of 0.74 mm; maximal fiducial distance was 0.81 mm. The mean distance of the 6 target points, the TRE, in the mouthpiece was 0.15 mm; maximal distance was 0.26 mm. Surfaces generated as described earlier resulted in 422,580 triangles before sterilization and 378,638 triangles after sterilization. The 2 directed Hausdorff
Table 1. HAUSDORFF DISTANCES AND FIDUCIAL AND TARGET REGISTRATION ERRORS BETWEEN COMPUTED TOMOGRAPHIC SURFACES GENERATED BEFORE AND AFTER STERILIZATION
Hausdorff distance After vs before Before vs after FRE (8 points) TRE (6 points)
2.51 2.42 0.81 0.15
mm mm mm mm
Abbreviations: FRE, fiducial registration error; TRE, target registration error. Figl et al. Hygiene and Deformation of Navigation Splints. J Oral Maxillofac Surg 2012.
HYGIENE AND DEFORMATION OF NAVIGATION SPLINTS
Distance Distribution Function 100
]%[ stniop 70 60 50 40 30 20 10 0
FIGURE 3. Distribution function of point-to-surface distances on the computed tomographic surface before and after sterilization; 90% of point distances are shorter than 0.61 mm. Figl et al. Hygiene and Deformation of Navigation Splints. J Oral Maxillofac Surg 2012.
distances were 2.42 mm for surfaces after versus before and 2.51 mm for surfaces before versus after sterilization. Symmetric Hausdorff distance, describing the maximal deformation of the splint, therefore was 2.51 mm. Results are presented in Table 1. The distribution function of the point-to-surface distances computed to obtain the Hausdorff distance after
versus before sterilization is displayed in Figure 3, which shows that 90% of values were below 0.61 mm and 95% were below 0.73 mm. A color-coded image of the splint is shown in Figure 4; the colors of the splint represent the distances of the appropriate point on 1 surface to the registered surface.
FIGURE 4. Left, Color-coded point-to-surface distances after versus before sterilization. Right, Surface before sterilization. A key to the colors is presented (far left). Figl et al. Hygiene and Deformation of Navigation Splints. J Oral Maxillofac Surg 2012.
FIGL ET AL
None of the indicators showed bacterial growth after the sterilization process.
Discussion Medical devices that have contact with sterile body tissues or fluids are considered critical items. These items, such as a splint used in a sterile operation field, must be sterile when used because any microbial contamination can result in infection. If these items are resistant to heat, the recommended sterilization process is steam sterilization because it has the largest margin of safety. However, reprocessing heat- and moisture-sensitive items such as splints requires a low-temperature sterilization technology. The present results showed that even after 5 lowtemperature sterilization cycles, the shape of the splint was altered to a very limited extent. In consequence, deformation after 1 sterilization cycle sufficient for clinical practice can be expected to be even smaller. As shown in Figure 4, longer distances occurred in the bent region next to where the mouthpiece was mounted. This might have been caused by tensions during the manufacturing process of the splint, suggesting some relevance of its design and fabrication, although the splint is not exposed to high temperature during sterilization. Although the fiducial registration error is an oftenused measurement quantifying registration accuracy, it depends very much on the splint-specific localization of the fiducial markers. A simple distance measurement between 2 locations of the same target markers can be used to estimate the effects within the target area, whereas the Hausdorff distance is a good measurement for the maximal overall deformation of the splint. By definition, the Hausdorff distance is very sensitive to outliers; therefore, the 95% percentile was used instead to provide more information. Alternatives to Sterrad sterilization in the present situation are rare. Because acrylic splints are very sensitive to temperature, they cannot be exposed to 121°C or 134°C as needed for steam sterilization, and the long aeration time after ethylene oxide treatment is unacceptable. Deformation-caused TRE and point-surface distance errors as measured using the Hausdorff distance will also depend on the size of the splint. Although these errors most likely will be smaller for smaller splints as they are used,16 registration accuracy increases with the size of the configuration.13 As an alternative registration method, bone screws have been used in dental procedures; Birkfellner et al17 used titanium microscrews for registration. Especially in patients with significantly atrophic alveolar ridges, this method has superior accuracy. Neverthe-
less, splints are the far less invasive technique and, therefore, very important in maxillofacial surgery.2 This study indicates that using hydrogen peroxide– based low-temperature plasma sterilization is a feasible method of reprocessing heath-sensitive acrylic navigation splints. Most importantly, the method does not deform the splint, which is the base for correct computer-navigated surgery. Acknowledgment This publication was supported by the Austrian Research Promotion Agency FFG (grant 818041B1), proposed and managed by Dr Schicho. Dr Weber and Dr Figl were partly supported by FWF project P19931. A preliminary version of this study was part of a poster presentation by Christoph Weber et al at the SPIE Medical Imaging Conference, San Diego, 2009.
References 1. Schicho K, Figl M, Seemann R, et al: Comparison of laser surface scanning and fiducial marker-based registration in frameless stereotaxy. Technical note. J Neurosurg 106:704, 2007 2. Fenlon MR, Jusczyzck AS, Edwards PJ, et al: Locking acrylic resin dental stent for image-guided surgery. J Prosthet Dent 83:482, 2000 3. Eggers G, Mühlinga J, Marmulla R: Image-to-patient registration techniques in head surgery. Int J Oral Maxillofac Surg 35:1081, 2006 4. Ewers R, Schicho K, Wagner A, et al: Seven years of clinical experience with teleconsultation in craniomaxillofacial surgery. J Oral Maxillofac Surg 63:1447, 2005 5. Klug C, Schicho K, Ploder O, et al: Point-to-point computerassisted navigation for precise transfer of planned zygoma osteotomies from the stereolithographic model into reality. J Oral Maxillofac Surg 64:550, 2006 6. Schicho K, Figl M, Seemann R, et al: Accuracy of treatment planning based on stereolithography in computer assisted surgery. Med Phys 33:3408, 2006 7. Ewers R, Schicho K, Undt G, et al: Basic research and 12 years of clinical experience in computer-assisted navigation technology: A review. Int J Oral Maxillofac Surg 34:1, 2005 8. Marcos-Martin M, Bardat A, Schmitthaeusler R, et al: Sterilization by vapour condensation. Pharm Technol Eur 8:24, 1996 9. Schaffer S, Pflug I: Vaporized hydrogen peroxide at low pressures as an agent to kill bacterial spores. Zent Steril Cent Serv 8:198, 2000 10. Lide D: CRC Handbook of Chemistry and Physics. Boca Raton, FL, CRC Press, 1994 11. Addy T: Low Temperature Plasma: A New Sterilization Technology for Hospital Applications. Morin Heights, Quebec, Canada, Polyscience Publishing Group, 1991 12. Horn BKP: Closed form solution of absolute orientation using unit quaternions. J Opt Soc Am A 4:629, 1987 13. Fitzpatrick JM, West JB, Maurer CR Jr: Predicting error in rigid-body point-based registration. IEEE Trans Med Imaging 17:694, 1998 14. Villard P, Bourne W, Bello F: Modelling organ deformation using mass-springs and tensional integrity. ISBMS 5104:221, 2008 15. Aspert N, Santa-Cruz D, Ebrahimi T: MESH: Measuring error between surfaces using the Hausdorff distance. Proc IEEE ICME 705:1, 2002 16. Gellrich NC, Schramm A, Hammer B, et al: Computer-assisted secondary reconstruction of unilateral posttraumatic orbital deformity. Plast Reconstr Surg 110:1417, 2002 17. Birkfellner W, Solar P, Gahleitner A, et al: In-vitro assessment of a registration protocol for image guided implant dentistry. Clin Oral Implants Res 12:69, 2001