Учебники / Middle Ear Mechanics in Research and Otology Huber 2006

FROM IMAGING OF THE MASTOID TO MECHATRONIC SURGERY: THE ROBIN-PROJECT

Marcus M. Maassen1, Florian Damman3, Jesus Rodriguez Jorge1, Dirk Malthan2, Armin Schäfer2, Erwin Schwaderer3, Raphael Ciuman1, Jan Stallkamp2,

Stefan Wössner2, Dirk Bartz5, Zein Salah4, Hans P. Zenner1

1 Laboratory for Medical Robotics, Department of Otolaryngology, University Hospital of Tuebingen, Germany

2 Fraunhofer-Institute for Manufacturing Engineering and Automation (IPA), Stuttgart, Germany

3 Department of Radiology, University Hospital of Tuebingen, Germany

4 WSI/GRIS – VCM, University of Tuebingen, Germany

5 ICCAS/VCM, University of Leipzig, Germany

For the past years, surgeons performing lateral skull base surgery have been paying more attention to the development and advancement of mechatronically aided surgical procedures. Especially the interconnection between electronic hearing implants (i.e. MET Otologics Transducer) and the ossicular chain requires an accurate milling of skull cavities to ensure a proper angle of coupling and a long term stable fixation. We developed a PC-based computer and robot assisted system (CAS/RAS) for the precise ande cientplanningandexecutionofsurgicalprocessesatthelateralskull.Employing multimodal sensory feedback, the geometry of the surgical area can be analyzed with a resolution of several µm.

158

1. Introduction

In otosurgery, one of the most challenging surgical procedures is the implantation of electronic hearing devices, requiring a highly reproducible andaccuratefixationtechnique.Here,theinterconnectionbetweentheam- plifier’svibratoryactorandtheossicularchain,andthefixationofso-called mountain brackets, requires an accurate milling of the required skull cavities [1]. In order to avoid a distortion of the transmitted acoustical signals, the implantation angle has to be defined very carefully. In addition, milling of the implant bed for the electronic module is usually time-consuming. Mechatronical devices may have the potential to perform this milling in a

by a modified industrial robot system “Adept Six-300” (Adept Technology GmbH, Dortmund, Germany). The employed robot system distinguishes six degrees of freedom (6-DOF) of its articulated kinematics allowing for flexible preparation angles. It is equipped with a 6-DOF force and momentum sensor system “Adept Force” (Adept Technology GmbH, Dortmund, Germany) so that any impact on the robot’s tool tip can be recorded and supervised.

In addition to the mechanical sensor systems, two di erent optical sensor systems have been utilized. A laser-based triangulation distance sensor “LT 100” (ElcoTec GmbH, Dortmund, Germany) has been integrated for the precise identification of the in-situ geometry. It consists of a laser source being reflected by the patient’s surface and a CCD-camera recording the reflection angle. This way, in combination with the robot’s kinematics, a laser scanner has been realized, allowing a determination of the patient’s topology with a precision better than 10 m in every direction. For the di erentiation of tissues, an auto fluorescence sensor system “S2000”, (OceanOptics Europe B.V., Duiven, The Netherlands) has been integrated. The tissue region to be analysed is radiated with ultra-violet light, the resulting auto fluorescence spectra are recorded and automatically interpreted. The milling of the skull cavities is done by a conventional bone drilling system “Elan EC” (Aesculap AG & Co KG, Tuttlingen). This way, the results of the robot-assisted intervention are comparable to the manual preparation of the required skull cavities.

4. Closed-loop Robot Control

The integration of sensory systems feeding back in-situ information allows foranemploymentoftherobotsysteminaso-calledclosedloopcontrolar- chitecture. Here, actual and nominal values are compared continuously, so

160that e.g. planned milling paths can be adjusted to intra-operative changes. This results in an increased precision and safety of the milling procedure.

The surveillance of the a ecting forces assures non-pathologic force impacts on the patient. The measured forces are recorded and transmitted to the control system. For the estimation of forces within a surgical step, algorithmswereimplemented,thatallowasimulationoftheforceimpacts. In case of unacceptable errors, the robot-assisted surgical process is interrupted. Current research e orts show that the analysis of recorded forces may help identifying anatomical structures e.g. dura mater [8]. These results may supplement the developed sensory system in the nearer future.

5. Topology Comparison and Registration

In particular the transition of the computer-assisted planning of a surgical proceduretoitsexecutionisregardedasacriticalstep.Forthisstepatransformation of the planning coordinate system into that of the patient is necessary: the patient has to be registered. Within the developed system, the comparison of the nominal geometry against the actual surface geometry using a laser scanner builds the basis for an intra-operative and continuous registration step.

In contrast to existing surface-based registration techniques like ICP- (Iterative Closest-Points) algorithms, the developed algorithms do not match the whole surface to realize a registration. Rather, significant associated anatomical landmarks in the actual and nominal data set are identified and assigned to one another. The technique is based upon a procedure called “spin-images” [9], was however modified in important pre and post processing steps. The advantages include the robustness against malformed surface reconstructions, resulting e.g. from blood and rinsing liquids.

The precision of the surface-based registration was evaluated by the measuring of artificial landmarks in form of registration metal beads. After the performed surface-based automatic registration, the position of the registration beads was identified by the robot system. For the determination of the robustness of the procedure a test series with 50 experiments was performed (Fig. 1), whereby the head model was deformed using polygon reduction and smoothing, and the addition of error noises [10]. The registration was accomplished in 48 of the 50 cases with an absolute error smaller than +/–0.7 mm (Fig. 2). In two cases the registration was, however, not accomplished correctly.

Furthermore, within the milling process the planned and simulated surface topology of the skull cavity can be compared with the performed milling process. Prior to the shift of the milling head, the in-situ geometry 161 is evaluated in real-time. Only in case of negligible error values, the mill-

ing process is continued. In other cases, a re-registration of the patient is performed automatically.

Fig. 1 Laboratory setup for milling tests on animal samples.

Fig. 2 Visualization of milling error by comparison of simulated and actual surface.

6. Conclusion

A computer and robot assisted system for microsurgical high-precision interventionsatthelateralskullbasehasbeendeveloped.Thesystememploys di erent sensor systems, like force sensors and laser triangulation sensors

162forprecisionmeasurements,inordertorealizequalityassuranceanddocumentation steps. Employing multimodal sensory feedback, the geometry of the surgical area can be analyzed with a resolution of several µm. Using this enhanced data, a robust surface based registration procedure could be implemented. Its evaluation showed a success rate of 96% with an absolute error smaller than +/–0.7 mm. It could be shown that the feedback of sensory information using spin image technology improves the safety and the precision of mechatronical-assisted surgical procedures [11].

ThisworkwassupportedbygrantNo.MA1458/2oftheGermanResearchFoundation (DFG).

References

1.Maassen M.M., Dammann F., Malthan D., Stallkamp J., Schwaderer E., Zenner H.P.: Development of a robot prototype with sensory feedback. 3rd Symposium of Middle Ear Mechanics in Research and Otology. Matsuyama, Ehime, Japan, 9th–12th July, 2003

2.Dammann F., Schwaderer E., Seemann M.: Einsatz von Methoden der 3D-Visua- lisierung im Kopf-Hals-Bereich. Fortschr Röntgenstr 2003; 175: 86.

3.Plinkert P.K., Plinkert B., Hiller A., Stallkamp J.: Einsatz eines Roboters an der lateralen Schädelbasis. Evaluation einer robotergesteuerten Mastoidektomie am anatomischen Präparat. HNO. 2001 Jul; 49 (7): 514–522

4.http://www.vtk.org, 27.02.2004

5.http://www.itk.org, 27.02.2004

6.http://xml.apache.org/xerces-c/index.html, 27.02.2004

7.Bowen, J., Stavridou V.: Safety-critical systems, formal methods and standards. Oxford University Technical Report PRG-TR-5-92, Software Engineering Journal, 178–183

8.Federspil P.A., Geistho U.W., Henrich D., Plinkert P.K.: Development of the First Force-Controlled Robot for Otoneurosurgery. Laryngoscope 2003 Mar; 113 (3): 465–471

9.Johnson, A.E.: Spin-Images: A Representation for 3-D Surface Matching. Carnegie Mellon University, Pittsbugh, Pennsylvania, August 1998

10. Malthan D., Stallkamp J., Wössner S., Dammann F., Zenner H.P., Maassen M.M.:

Lasertriangulationsgesteuerte Sensorik und miniaturisierte Roboterkinematik können Implantationen an der lateralen Schädelbasis verbessern, 5. Oktober 2002, Leipzig, II Jahrestagung der Sektion Neuroendoskopie, Neuronavigation und intraoperative Bildgebung der Deutschen Gesellschaft für Neurochirurgie.

11. Malthan D., Ehrlich G., Stallkamp J., Dammann F., Schwaderer E., Maassen M.M.: Automated registration of partially defective surfaces by local landmark identification. Comput Aided Surg. 2003; 8 (6): 300–309

163

OSSICULAR RECONSTRUCTION USING STORED INCUS IN PLANNED TWO-STAGE TYMPANOPLASTY

Kiyofumi Gyo, Toshiki Maetani, Naoto Hato, Masamitsu Hyodo

Address: Kiyofumi Gyo, MD, Department of Otolaryngology, Ehime University Shizukawa, Toon, Ehime 791-0295, Japan, Email: kiyofumi@m.ehime-u.ac.jp

Keywords: ossicular chain, mastoid bowl, cholesteatoma, staged tympanoplasty,

The incus once removed at the first stage can be used in ossicular reconstruction at the second stage, when it is preserved appropriately. The present study was conducted to investigate if the mastoid bowl is a suitable site for storage. The study group included 24 ears of 23 patients who underwent staged tympanoplasty for the treatment of middle ear cholesteatoma. The average interval between the two stages was 8.3 months (range 6–12 months). The incus was identified in all cases at the second stage: ten incudes were found to be covered with a thin mucosa layer, 12 were buried in fibrous or granulation tissue, and two were joined to the surrounding bone. In 19 cases, the incus was available as a short columella for ossicular reconstruction. The remaining five cases were reconstructed using a hydroxyapatite ossicle as a long columella, since the stapes superstructure was missing at the second stage. Preservation of the incus

164in the mastoid bowl is an e ective option in staged tympanoplasty, when the incus is considered useful for ossicular reconstruction at the second stage.

1. Introduction

According to Sheehy and Crabtree [1], planned two-stage tympanoplasty is indicated when the middle ear is severely a ected by diseases such as middleearcholesteatoma.Atthefirststage,theincudostapedialjointisdisarticulated and the incus is removed for safe, complete exenteration of the disease. When the incus is deemed useful as a columella in ossiculoplasty, it is often stored outside the body until use at the second stage. We previously kept the incus in a tiny cell filled with pure alcohol and stored it in a

refrigerator. This procedure is convenient for eradicating possible residual squamous epithelium attached to the incus. However, it also includes risks, such as accidental loss during storage, mistaken use as someone else’s ossicle, viral or bacterial contamination, and death of the living osteocytes, which may induce atrophy of the incus after repositioning.

To prevent these problems, we recently stored the incus in the eradicated mastoid bowl after stripping the diseased surface, when the incus seemed useful for ossicular reconstruction. This paper describes the results of stored incus at the second stage and the hearing outcomes of ossiculoplasty by use of it.

2. Methods

Subjects were 24 ears of 23 patients who underwent staged tympanoplasty fortreatingextensivemiddleearcholesteatoma.Thepatientswere12males and 11 females, and their ages ranged from 4 to 62 years, with an average of 25.6 years. The patients were chosen for this study because the incus remained in good shape and was considered suitable as a columella in ossicularreconstructionatthesecondstage.Surgerywasconductedusingthe intact canal wall technique.

Resultsofossiculoplastywereevaluatedusingthecriteriaproposedby the Otological Society of Japan in 2000. Audiometric data were obtained by calculating the pure-tone averages of 500, 1000, and 2000 Hz before the first stage, and at least 6 months after the second stage. The surgery was considered successful when at least one of the following three criteria was achieved: a postoperative air-bone gap within 15 dB, a hearing gain of more than 15 dB, or a threshold of postoperative air conduction hearing within 30dB.

165

3. Results

At the second stage, the incus was identified at the original sites of storage inallcases;10incudeswerefoundtobecoveredwithathinlayerofmucosa (Fig. 1), 12 were buried in fibrous or granulation tissue (Fig. 2), and 2 had adhered to the surrounding bone via bony adhesions (Fig. 3). These results were closely related to the preoperative condition of the mastoid cavity; when the mastoid disease was mild and the mucosa was preserved, the incus tended to be covered with a thin layer of mucosa. When the mastoid cavitywasseverelydiseased,necessitatingextensivemastoidectomy,theincustendedtobeburiedinfibrousorgranulationtissue.Ageandgenderdid not influence the outcomes of the incus.

Residual cholesteatoma was discovered as a pearly mass in 6 ears, either at the attic (3 ears) or the tympanic sinus (3 ears). These cholesteatomas were localized and easily removed. No residual disease was ever observed in the mastoid bowl where the incus had been stored.

In19of24cases,theossicularchainwasreconstructedusingtheincus as a short columella. In the remaining 5 cases, the stapes superstructures thathadbeenidentifiedatthefirststagedisappearedbetweenstages.These 5 cases were reconstructed using a hydroxyapatite ossicular prosthesis as a long columella.

Postoperative hearing was assessed in 17 of the 19 patients who had undergone ossicular reconstruction with a stored incus; 2 cases were lost for follow-up. When the latest audiometric results were analyzed, the ossiculoplasty was judged success in 11 out of 17 cases (65%); there was closure of the air-bone gap within 15 dB in 11 cases (65%), a hearing gain of more than 15 dB in 5 cases (18%), and postoperative air conduction hearing within 30 dB in 9 cases (53%).

4. Discussion

Although ossiculoplasty is performed using a variety of materials, autologous ossicles are considered best as columellae because of their easy availability,lowcostofpreparation,andgoodbiocompatibility.Theyalsoresulted in good postoperative hearing in a long-term follow-up study. At present, however, use of incus in the cholesteatomatous ear remains controversial because of the possibility of recurrent cholesteatoma and progressive bone absorption. Steinbach et al. [2] studied 203 incudes taken from patients with cholesteatoma. Their histological examination revealed the frequent existence of tiny deposits of squamous tissue on the incudes that appeared free of cholesteatoma under an operating microscope. Dornho er et al.[3]

also stated that gross examination of the incus with an operating micro- 167 scope was not accurate in assessing cholesteatoma involvement. To reduce

the risk of cholesteatoma recurrence, Muller-Hermann [4] advocated that the incus be autoclaved during surgery to kill the squamous epithelium adhering to it, although the procedure simultaneously destroys the living osteocytes.

Inthecaseofstagedtympanoplasty,wecanavoidtheproblemsrelated to reuse of the incus. The simplest procedure is to store the incus in a cell filledwithalcoholuntilitisusedatthesecondstage.Thisprocedureissufficient to kill any squamous epithelium anchored to the incus, although it was somewhat cumbersome and risky. Storage of the incus in the mastoid bowl during staging can overcome most of the disadvantages of external