Full text of DOI:10.1016/S0194-5998(00)70006-4

Frameless computer-aided surgery system for revision
endoscopic sinus surgery
MARCO CAVERSACCIO, MD, RICHARD BÄCHLER, MSc, KURT LÄDRACH, MD, DMD, GERHARD SCHROTH, MD,
LUTZ-PETER NOLTE, PhD, and RUDOLF HÄUSLER, MD, Bern, Switzerland
1
To increase the intraoperative safety factor and to
acquire anatomic assistance during revision endo-
scopic sinus surgery (RESS), we used an optical
computer-aided surgery (CAS) system that we
developed collaboratively in Bern, Switzerland.
During 1 year, 25 RESSs were performed with CAS:
recurrent polyposis (n = 20), recurrent frontal recess
stenosis (n = 3), and recurrent frontal recess steno-
sis with mucocele (n = 2). These patients were
compared with a control group of 10 patients
undergoing RESS without CAS. The same surgeon
computer-aided surgery (CAS). Different systems can
be used to localize an instrument: electromechanical
2
3
4
systems, electromagnetic systems, optical systems,
5
or ultrasonography. Each system uses a different prin-
ciple to locate the exact position of a marked surgical
instrument that is followed online on the CT or MRI
images during surgical intervention. These systems
were first conceived for use during neurosurgical proce-
6
dures in which the head was rigidly fixed. Fixation of
the head is not suitable for endoscopic paranasal sinus
surgery because the head must often be moved to obtain
access to deep-seated structures. Therefore a novel
frameless CAS system has been developed at the
Maurice-E. Müller Institute for Biomechanics in Bern,
in collaboration with the University Clinic of ENT,
Head and Neck Surgery at the Inselspital. The system
is now used routinely for anterior and lateral skull base
surgery and has proven to be a success. In this article
(
M.C.) performed all operations, and there were no
minor or major complications in either group. The
clinical inaccuracy of our system is between 0.5
and 2 mm with paired-point and surface matching.
The navigation system is an important aid to sur-
geons in identifying anatomic landmarks that are
typically difficult to visualize in this type of surgery,
thus reducing the stress placed on the surgeon.
7
8
we report the results we have obtained by using the
Bernese CAS system for revision endoscopic sinus
surgery (RESS) over a 1-year period.
(Otolaryngol Head Neck Surg 2000;122:808-13.)
Our goals were to evaluate how much time is neces-
sary for this procedure and to determine whether the
system is an aid in identifying anatomic structures in a
complex area without important missing landmarks.
M
odern-day imaging techniques, such as CT and MRI,
permit detailed identification of very fine anatomic struc-
tures. Their primary clinical use traditionally has been for
the topographic diagnosis of illness and injuries.
However, an application that has begun to emerge in the
last few years is the use of these imaging techniques for
METHODS
Informed consent was obtained from patients, but the
study was classified as exempt by the local institutional
review board.
From the Department of Otorhinolaryngology–Head and Neck
surgery (Drs Caversaccio, Lädrach, and Häusler) and the
Department of Neuroradiology (Dr Schroth), Inselspital, University
of Bern; and M.-E. Müller Institute for Biomechanics (Mr Bächler
and Dr Nolte), University of Bern.
Supported by a grant from the Department of Clinical Research,
University of Bern.
Presented in part at the 1st International Congress on Computer
Integrated Surgery in the Areas of Head and Spine, Linz, Austria,
September 1-5, 1997.
Characteristics of the Optical CAS System
The concept developed by the research group in Bern is
based on CT scans. The images are saved in a digital archiv-
ing system by means of a network or directly transferred to the
computer in the operating room (Sun WorkStation [Unix],
Sun Microsystems). A frameless optical system is used in the
operating room. Intraoperative digitization through an infra-
red system is based up the 3-dimensional motion analysis sys-
tem (Optotrak 3020, Ontario, Canada; Fig 1). The camera bar
contains 3 one-dimensional sensors, each of which uses a
Reprint requests: Marco Caversaccio, MD, University Hospital,
Department of ENT, Head and Neck Surgery, Freiburgstrasse, CH-
3010 Bern, Switzerland.
Copyright © 2000 by the American Academy of Otolaryngology–
Head and Neck Surgery Foundation, Inc.
2
048 linear charge-coupled device array and a cylindrical
lens. It can track up to 256 pulsed infrared light-emitting
diode (LED) markers, with a maximum rate of about 3500
0
194-5998/2000/$12.00 + 0 23/77/99667
doi:10.1067/mhn.2000.99667
808

Otolaryngology–
Head and Neck Surgery
Volume 122 Number 6
CAVERSACCIO et al 809
Fig 2. The upper jaw (maxilla) splint with tooth impression
in the siloxane mass compound (after postoperative
removal) serves as the dynamic reference base. The rod
is used to affix the infrared LEDs.
Fig 1. Intervention with CAS for RESS. Visible is the infrared
camera (arrow), the workstation, and the position of the
surgeon.
markers per second. Within a field of view of 1.0 × 1.2 m at a
distance of 2.0 m, the camera can locate each LED with an
accuracy of 0.1 mm in the x and y axes and 0.15 mm in the z
axis, with a resolution of better than 0.01 mm. By attaching at
least 3 markers onto each rigid tool, the intraoperative local-
ization and orientation in space can be calculated trigonomet-
rically.
Operation Planning
If pathology in the sinuses must be surgically addressed,
an axial CT scan of the skull is made, with a sectional thick-
ness of 1.5 mm and a spacing of 1.5 mm or a helical scan.
Reliable paired points are then marked on the CT scan, such
as, in case of the anterior skull base, the frontozygomatic
sutures, the nasion, and the nasal spine. Neither fixation of the
head with frames nor insertion of invasive markers is needed.
In addition, possible operation trajectories (paths to the target
and goals) can be identified. These are especially relevant
with regard to the anatomic structures that will be encountered
and bypassed during the intervention.
Fig 3. The virtual keyboard, with its various functions,
allows the surgeon to carry out commands by way of the
computer workstation.
Operation
After the administration of the local anaesthetic, a dynamic
reference base is positioned on the head of the intubated
patient. It consists of an upper jaw (maxilla) splint equipped
with a rod and is attached to the jaw by means of a siloxane
impression material of medium consistency (Coltène, Alt-
stätten, Switzerland; Fig 2). The rod is equipped with LEDs.
If the patient is edentulous, the same splint with siloxane mass
is needed. Additionally, we fix the splint with 1 to 2 mini-
screws (diameter of 2.0 mm and length of 12-14 mm;
Leibinger, Freiburg, Germany) on the hard palate. The mini-
screws are left on until the end of the operation. This system
allows free head movement during the intervention. The sur-
geon controls the various computer functions by using a spe-
cially developed virtual keyboard (Fig 3). To control variabil-
ity in the study, all procedures were performed by the same
surgeon (M.C.) and compared with a control group of 10
patients undergoing RESS without CAS.
CAS Instruments and Evaluation of Instrument
Accuracy (Matching)
To use the tools with the CAS system, a calibration is
mandatory. Each tool is checked preoperatively for its accuracy

Otolaryngology–
Head and Neck Surgery
June 2000
810 CAVERSACCIO et al
Fig 6. Hopkins-Optic 30° endoscope (Storz ) equipped
with a rotatable 360° LED shield (90° step).
Nephew, Andover, MA]; Fig 5) and the endoscope (Hopkins-
Optic, Storz, Tuttlingen, Germany) are followed online on the
computer monitor.
Fig 4. The calibration unit is equipped with an LED shield
and with different holes for the tools (different sizes of
shells are at the surgeon’s disposal).
A 4-mm diameter endoscope equipped with 4 LEDs and
o
the possibility to turn 90 (Fig 6) is used for this purpose,
which provides the position of the endoscope on the monitor.
During surgery, the operator faces the computer workstation
monitor and observes the CT projections (axial, sagittal, and
coronal) and the endoscopically navigated picture. The setup
allows direct surgery by way of the monitor (Figs 1 and 7).
Patients
During 1 year, 25 RESSs were performed: recurrent poly-
posis (n = 20), recurrent frontal recess stenosis (n = 3), and
recurrent frontal recess with mucocele (n = 2). The subjective
advantages for the surgeon using the CAS were analyzed (Fig
8
A and B). A control group (n = 10; n = 8 with recurrent poly-
posis nasi and n = 2 with recurrent frontal recess stenosis)
operated on without CAS was compared with the group of 25
patients undergoing RESS with CAS.
Fig 5. Microdebrider (Dyonics, Smith Nephew) equipped
with LED.
RESULTS
We found clinically average inaccuracies at the ante-
rior skull base and the lamina papyracea between 0.5
and 2 mm (average, <1 mm).
in a calibration unit equipped with LEDs and different sizes of
holes (Fig 4). The calibration remains valid as long as the
marker probe is rigidly attached to the tool and no deforma-
tion of the tool has occurred. The single calibration unit used
allows for fast and secure calibration of all tools.
The actual position of the skull is compared with the CT
scan by using a needle pointer. For this purpose, the anatomic
paired points (landmarks) established the previous day and, if
necessary, points on the surface of the skull are used. By using
a computer algorithm that was developed specifically for this
system, these anatomic points are precisely correlated with the
CT scan (matching).
Before the operation begins, an anatomic check is carried
out in which known anatomic structures are evaluated online
for positional accuracy and to estimate the target error. During
the operation, the tips of the surgical instruments (eg,
Rhinoforce [Storz, Tuttlingen, Germany] and Dyonics [Smith
Although the operations were difficult in all cases,
there were no minor or major complications. The splint
was removed without any dental fracture, and in cases
with edentulous patients (n = 3), no bleeding or infec-
tion in the hard palate occurred. There were also no
complications in the control group of 10 patients (n = 8
for recurrent polyposis and n = 2 for recurrent frontal
recess stenosis) undergoing RESS without CAS.
The supplementary installation time for CAS was
always between 15 and 20 minutes in the 25 patients
undergoing RESS with CAS.
The operation time (same surgeon) with CAS or
without CAS depended on whether a radical or a func-
tional endoscopic revision surgery in recurrent polypo-
sis was made. In addition, the amount of time consumed

Otolaryngology–
Head and Neck Surgery
Volume 122 Number 6
CAVERSACCIO et al 811
depended on which technique was used in recurrent
9
frontal recess stenosis (Type Draf I-III). Because of the
small number of patients and the large parameters, sta-
tistical determination of whether the time spent during
the operation was less with CAS than in our control
group of 10 patients could not be determined.
The CAS subjectively gives the surgeon more secu-
rity and an anatomic aid during RESS in situations such
as when fibrous or granulomatous scar tissue without
anatomic landmarks is found (eg, uncinate process and
bulla ethmoidalis). In addition, it allows preview and
measurement of the distance to the bone border on the
CT slices.
DISCUSSION
Interventions at the skull base during revision opera-
tions are often difficult because of the presence of com-
plex anatomic structures that are difficult to identify.
According to the literature, minor or major complica-
tions can occur (eg, loss of vision and intracerebral neu-
rologic complications) as a result of surgical procedures
10
in this location. The use of computer-aided technology
can represent an important safety factor for this area of
surgery. The clinical accuracy achieved for our system,
as determined for 25 patients who underwent surgery of
the anterior skull base, was between 0.5 and 2 mm
(
average, <1 mm). The accuracy of other systems on the
Fig 7. Endonasal intervention with LED-equipped
Rhinoforce (Storz; right hand) and the endoscope (left
hand) with CAS. An LED-equipped splint (bottom right) is
fastened onto the patient’s upper jaw, which allows an
exact dynamic online display of the location of the tips of
the instrument, while still allowing free head movement.
11-13
market lies between 0.5 and 3 mm.
It is worth not-
ing that a scattering of the results (of the inaccuracies)
can be present and that the systematic and repeatable
reproducibility can be progressively improved up to 1
mm by pair-point matching and surface matching with
our system.
A clear advantage of the system we developed is that
it is frameless and allows unimpeded head movement.
The first use of frameless noninvasive stereotactic inter-
vention in neurosurgery with an upper jaw splint is
The time for the installation of the computer, the
infrared camera, the upper jaw splint, the instrument
calibration, and matching (comparison of the actual
positions with the previously taken CT) processes takes
15 to 20 minutes. During this time, as is the usual prac-
1
4
credited to Carini et al in 1992. The Bern research
group wanted no additional preoperative marking, such
as insertion of screws, fiducial markers, a headset, or a
4
,16
tice with other such systems,
the local anaesthetic is
administered in the operation field.
3
,11,15
fixed splint.
Therefore our system uses only one
Operating times are initially longer because of the
time needed to learn to work with a navigation system
and video-controlled endoscopy; however, after the first
few operations, the surgeon finds his rhythm, and
surgery can progress at an even pace. In our study we
did not find that CAS reduced the time needed for
RESS when compared with a control group of 10
patients who underwent RESS without CAS.
anchored upper jaw splint during the operation. It is
made of plastic with a metal rod and is attached to the
jaw with a siloxane impression compound (Coltène).
Additionally, in edentulous patient 1 or 2, miniscrews
(Leibinger, Freiburg, Germany) are needed to hold the
splint in place. This allows a dynamic reference base to
be affixed to the skull precisely.
In ear, nose, and throat surgery the patient’s CT scan
is stored in an archiving system, such as an optical disk
or a digital audiotape, or in a picture archiving system
so that no additional exposure to radiation is necessary,
and time used is reduced for the patient.
The anatomic aid provided by the CAS system for
RESS is very important for the surgeon. It provides
more security and reduces stress, which makes the oper-
ation more efficient. In addition, it allows preview and
measurement of the distance to the bone border on the

Otolaryngology–
Head and Neck Surgery
June 2000
812 CAVERSACCIO et al
A
B
Fig 8. Intraoperative instrument tracking with CAS on the anterior skull base. A, CT of the paranasal
sinuses in a patient operated on 2 times before (Samter syndrome) is shown on the monitor: axial,
sagittal, and coronal views and 3-dimensional reconstruction (note the absence of the anatomic
landmarks). The cross-hair corresponds online to the tip of the microdebrider at the sphenoid entry
visible on the endoscopic picture. B, Monitor view (axial, sagittal, and coronal view) with endoscop-
ic picture of a recurrent polyposis nasi and a concha bullosa on the left side.
CT slices, which gives the surgeon the possibility to be
more radical. A prospective study will be necessary
with a CT control after the healing phase (8 weeks) to
conclude whether RESS with CAS is more radical or
efficient than RESS without CAS.
Our rhinoinstruments (Blakesely forceps, scraper,
optics [Storz, Tuttlingen, Germany], and microdebrider
and calibrated before the operation with a specially man-
ufactured calibration unit. Subjectively, we found that the
microdebrider equipped with LEDs gives more security
than a normal forceps because the microdebrider cut
away and did not drag the tissue. The accuracy on the
lateral wall (lamina papyracea) with the bent forceps or
the microdebrider equipped with LEDs (blade on the
side) is the same, but the forceps (equipped with or with-
1
7
[
Dyonics, Smith-Nephew]) were equipped with LEDs

Otolaryngology–
Head and Neck Surgery
Volume 122 Number 6
CAVERSACCIO et al 813
out LEDs) seem better on the top (lamina cribrosa)
because the opening is on the front of the tool.
The sterile touch pad or virtual keyboard that we
have developed is equipped with LEDs and allows the
surgeon to remain independent by inputting commands
directly into the computer without the danger of dis-
turbing the sterile field. Additional computer personnel
are not needed.
A disadvantage of the optical or infrared camera sys-
tem is that no obstacle can be tolerated between the
tool, dynamic reference base, and infrared camera.
The cost of our system (Surgigate ORL, medivision;
Stratec Medical, Oberdorf, Switzerland; CHF: 300,000
3. Fried MP, Kleefield J, Gopal H, et al. Image-guided endoscopic
surgery: results of accuracy and performance in a multicenter
clinical study using an elecromagnetic tracking system. Laryngo-
scope 1997;107:594-601.
4. Anon JB, Klimek L, Mosges R, et al. Computer-assisted endo-
scopic sinus surgery. An international review. Otolaryngol Clin
North Am 1997;30:389-401.
5
. Reinhardt HF, Horstmann GA, Gratzl O. Mikrochirurgische
Entfernung tiefliegender Gefässmissbildungen mit Hilfe der
Sonar-Stereometrie. Ultraschall Med 1991;12:80-4.
6
. Spiegel EA, Wycis HT, Marks MM, et al. Stereotaxic apparatus
for operations on the human brain. Science 1947;106:349-50.
. Nolte LP, Zamorano L, Visarius H, et al. Clinical evaluation of a
system for precision enhancement in spine surgery. Clin Bio-
mech 1995;10:293-303.
7
8
. Caversaccio M, Lädrach K, Häusler R, et al. Konzept eines rah-
menlosen bildinteraktiven Navigationssystems für die
Schädelbasis-, Nasen- und Nasennebenhöhlenchirurgie. Otorhino-
laryngol Nova 1997;7:121-6.
[US: $200,000]) is easily offset by its advantages. There
are higher levels of safety, which means fewer potential
medicolegal consequences in the future.
An additional advantage of a CAS system is its
potential as a teaching tool for residents, both during an
actual surgery and during the preoperative planning
process.
In conclusion, the CAS system that we have devel-
oped in Bern is an aid to surgeons because it is conve-
nient and provides anatomic orientation, which is not
available otherwise. This raises the level of safety for
interventions for RESS without increasing the time
needed for the surgery. Future studies are necessary to
demonstrate whether RESS with CAS is more efficient.
9
. Draf W, Weber R, Keerl R, et al. Aspekte zur Stirnhöhlen-
chirurgie. Teil 1: Die endonasale Stirnhöhlendrainage bei
entzündlichen Erkrankungen der Nasennebenhöhlen. HNO 1995;
4
3:352-7.
1
0. Hosemann W. Die endonasale Chirurgie der Nasennebenhöhlen-
Komplikationen. Eur Arch Otorhinolaryngol Suppl 1996;1:231-
43.
1. Hauser R, Westermann B, Probst R. Noninvasive tracking of
patient’s head movements during computer-assisted intranasal
microscopic surgery. Laryngoscope 1997;107:491-9.
12. Carrau RL, Snyderman CH, Curtin HD, et al. Computer-assisted
intraoperative navigation during skull base surgery. Am J Oto-
laryngol 1996;17:95-101.
3. Fernandez PM, Zamorano L, Nolte L, et al. Interactive image
guidance in skull base surgery using an optical device. Skull Base
Surg 1997;7:15-21.
4. Carini S, Calcagno E, Tortori-Donati P, et al. A new model for
non-invasive, reproducible fixation of a stereotaxic frame using
an orthodontic resin plate. Acta Neurochir 1992;118:159-61.
5. Bale RJ, Vogele M, Freysinger W, et al. Minimally invasive head
holder to improve the performance of frameless stereotactic
surgery. Laryngoscope 1997;107:373-7.
16. Freysinger W, Gunkel AR, Thumfart WF. Image-guided endo-
scopic ENT surgery. Eur Arch Otorhinolaryngol 1997;254:343-6.
17. King JM, Caldarelli DD, Pigato JB. A review of revision func-
tional endoscopic sinus surgery. Laryngoscope 1994;104:404-8.
1
1
1
1
We thank Erwin Häsler for photo assistance.
REFERENCES
1. Jolesz FA. Image-guided procedures and the operating room of
the future. Radiology 1997;204:601-12.
2. Mösges R, Schlöndorff G. A new imaging method for intraoper-
ative therapy control in skull-base surgery. Neurosurg Rev 1988;
1
1:245-7.