Acta Oto-Laryngologica. 2014; 134: 352–357
ORIGINAL ARTICLE
Magnetically driven middle ear ossicles with laser vibrometry as a new diagnostic tool to quantify ossicular fixation JOHN PEACOCK1, JORIS DIRCKX1 & MAGNUS VON UNGE2 1
Laboratory of Biomedical Physics, University of Antwerp, Belgium and 2Department of Otorhinolaryngology, Akershus University Hospital and University of Oslo, Oslo, Norway
Abstract Conclusion: Information on the degree of incus fixation can be gathered by measuring the ratio of incus to umbo long process velocity through the ear canal. Objectives: To test a new method of quantifying partial ossicular fixation in an ear with an elevated tympanic membrane. Methods: Measurements were made on four fresh-frozen human temporal bones. After elevating the tympanic membrane a small magnet was attached to the manubrium and an electromagnetic excitation coil was used to vibrate the ossicles. The vibration response of the tip of the incus long process and the umbo were measured before and after artificially fixating the incus to the lateral attic wall. Results: Partial incus fixation resulted in a decrease in both the incus and umbo velocities, with the incus velocity being more severely reduced. The decreased ratio of their vibrations is a clear indicator of the degree of incus fixation.
Keywords: Otitis media, ear surgery, incus fixation, umbo long process
Introduction A variety of conditions can result in a stiffening of the middle ear (ME) ossicles, which leads to hearing impairment. The ossicles may be more or less fixated due to various kinds of pathologies such as otosclerosis and malformations, and sequels of chronic otitis media such as tympanosclerosis and adhesions. In order to remedy a patient’s hearing loss surgical intervention is often required, and in order to determine the best course of surgical treatment some knowledge of the degree of ossicular mobility is useful. The final ossicular assessment is very often made by means of manual palpation during surgical exploration. Such assessment is difficult and the exact state of the ossicles can still remain unclear. Furthermore, manual pressing tests the ME’s quasi-static mechanical properties, which are not necessarily the same as the dynamic properties that govern sound transmission. A more objective method to assess mobility would be helpful.
A laser Doppler vibrometer (LDV) is a device that allows high precision noncontact vibration measurements on the nanometre scale. It has become a standard tool to measure ossicular motion, having been used to study ME mechanics in animals, human cadaveric temporal bones, and even in live human subjects [1–4]. Yet, despite the evident value of the technique, it has not found its place in routine clinical practice. Vibration measurements at different points along the ossicular chain can be made in temporal bones after drilling an opening in the facial recess. This method has been used to make measurements of stapes motion in live human subjects undergoing cochlear implantation [1,2]. However, in most operations this approach is too invasive, and for diagnostic purposes a means of measuring partial ossicular fixation through the ear canal would be preferable. During endaural surgery, the tympanic membrane (TM) is elevated giving free visual access to a large portion of the ossicular chain. However, in contrast to
Correspondence: Mr John Peacock, Laboratory of Biomedical Physics, University of Antwerp, Groenenborgerlaan 171, B-2020, Antwerp, Belgium. Tel: +32 (0) 3 265 34 38. E-mail: [email protected]
(Received 6 June 2013; accepted 23 August 2013) ISSN 0001-6489 print/ISSN 1651-2251 online 2014 Informa Healthcare DOI: 10.3109/00016489.2013.841990
Magnetically driven middle ear ossicles the situation above, where the TM can be driven by acoustic stimulation from the ear canal, the endaural surgery with elevation of the tympano-meatal flap prevents this approach. Some authors have indicated that measurements on the TM itself (before surgical exploration) can be used to evaluate the degree of ossicular fixation [3,4], but in many cases success is limited as partial ossicular fixation does not always result in reduced umbo velocities that are large enough to discriminate stiffened ossicles from normal ones, or to discriminate different types of pathology. Others have developed devices to measure partial fixation after elevating the TM [5], but this involves making physical contact with an actuator. The method has not gained widespread acceptance, possibly because special skill is required to make reliable mechanical contact between the actuator and the ossicles without overloading or hampering their motion. This study describes a new technique to measure ossicular vibration. The technique makes use of a small magnet and coil to vibrate the umbo after the TM is elevated and dissected free from the handle of the malleus, thus allowing for measurements of incus vibration through the ear canal. In this way data may be obtained in live humans with limited surgical intervention. The potential clinical usefulness of the technique is discussed. Material and methods Measurement set-up Four fresh frozen human temporal bones, acquired from Medcure Inc., Portland, Oregon, USA, were used. Before beginning any measurements they were left at room temperature for several hours to defrost, and were examined under an operating microscope for signs of abnormality. The measurement set-up is illustrated in Figure 1. A measurement set-up was built that consisted of a laser Doppler vibrometer (OFV-534, Polytec GmbH, Waldbronn, Germany) coupled to a surgical microscope (OPMI Sensera/S7; Carl Zeiss, Oberkochen, Germany). A small mirror, the orientation of which was controlled by piezo motors, was placed in front of the vibrometer, and could be used to accurately position the laser beam (spot size 200 mm at a distance of 30 cm). For comparison to stimulation with a magnet, an acoustic stimulus was provided by a free-field loudspeaker system, the output of which was measured by a probe microphone (Bruel & Kjær probe microphone, type 4182; Nærum, Denmark). The signal was designed in a PC using MATLAB and generated
353
Surgical microscope Vibrometer Temporal bone Speaker
Probe microphone
Coil
Figure 1. Drawing of the set-up. The dotted line shows the path of the laser beam.
using an A/D-D/A conversion board (RME HDSP 9632; Audio AG, Hainhausen, Germany) with 24-bit resolution and a sampling rate of up to 192 kHz. A stepped sine signal consisting of logarithmically spaced frequencies (with 16 per octave) between 250 and 8000 Hz was used. The sound pressure level (SPL) was adapted at each frequency to match an equal loudness curve at 80 phons (the relationship between loudness and SPL for humans is defined in ISO norm 226:2003). With an SPL equal for all frequencies, not all frequencies will sound equally loud to the human ear. By adapting the signal to the ear’s characteristics, it is possible to use higher stimulation levels at high and low frequencies without causing damage. An equal loudness level of 80 phons means that tones at all frequencies will sound, to the human ear, equally as loud as a 1000 Hz tone at 80 dB SPL. A small custom-made neodymium magnet (1 mm diameter 0.5 mm thickness, 0.003 g), in combination with a custom-built pancake coil (100 mm inner and 200 mm outer diameter, 97 spiral coiled windings) was used for the magnetic stimulation. The coil was driven by a power amplifier (DP P-900 Vintage Stereo Amplifier; Highlite International B.V., Kerkrade, the Netherlands) with the signal being constructed in MATLAB, the same as for the acoustic signal.
Measurement procedure To expose the ossicles and allow artificial fixation, the attic of the ME cleft was opened by drilling a hole in the tegmen tympani from the middle cranial fossa side. The approximate size of the hole was 5 mm. The
354
J. Peacock et al. 10
Velocity (mm/s)
1
TB1 TB2 TB3 TB4
Unfixed Fixed
0.1 0.01 0.001
500
1000
2000
4000
8000
Frequency (Hz) Figure 2. Umbo velocity before and after artificial fixation. The darker lines indicate the unfixed state and the lighter lines the fixed state. In temporal bone (TB) 1 and TB4 the velocity in the fixed state is in the same range as the unfixed condition for the other two bones.
hole gave access to the short process of the incus, the incus body, and the head of the malleus. Sound was presented to the TM, and the sound levels were calibrated using the probe microphone with its tip placed close to the opening of the ear canal. The calibrated sound was then used to stimulate the TM and data on the vibration response of the umbo were measured by the vibrometer. To improve the strength of the reflected vibrometer signal, a small piece of reflective tape (Polytec) was placed at the umbo. The TM was then elevated and pieces of reflective tape were placed on the umbo and at the tip of the long process of the incus. The magnet was then attached to the manubrium, at a point close to the umbo, using a small drop of glass ionomer luting cement (GC Corporation, Tokyo, Japan). It takes a few minutes to set and fixes the magnet quite rigidly. The custom-built pancake coil was placed over the ear, with the opening to the ear canal near its center, as if the coil was resting on a patient’s head. A signal was then put to the coil to vibrate the magnet, and the vibrometer was again used to measure the response of the umbo. The vibration response measured with magnetic stimulation was compared to that achieved with the earlier acoustic stimulation and the signal to the coil was adjusted, by means of an iterative procedure, until the umbo response matched that achieved acoustically to within 1 dB. Three separate measurements were made at each measurement point (three at the umbo and three at the tip of the long process of the incus). Subsequently, the incus was artificially fixed by the application of luting cement (the same kind as was used to attach the magnet) between its head and the wall of the attic bone. A drop of cement was applied to fill out the space between its most superior portions (the head and short process) and the lateral attic wall of the temporal bone. After the cement had set the
measurements were repeated as before. The time taken for each measurement was just under 6 s. Results All four temporal bones were examined under an operating microscope. The TM was examined and appeared normal, and after TM elevation the ME showed no signs of abnormality. The mucosa was examined for adhesions or scar tissues and the ossicles appeared normal. No other signs of pathology were detected. Figure 2 shows the velocity of the umbo before and after partial incus fixation for all four temporal bones (labelled TB1–4). In TB2 and TB3 the vibration response clearly diminishes after the cement is applied. In TB1 and TB4 the response does drop, as compared with the original vibration level in each temporal bone, but the vibration level after fixation in these specimens does not differ significantly from the unfixed vibration level. Figure 3 shows the velocity of the incus long process before and after partial incus fixation for all four temporal bones. In contrast to measurements on the umbo, the response of all temporal bones clearly diminishes after applying the cement and the change in response is larger than for the umbo. Figure 4 shows the ratio of incus long process to umbo velocity before and after fixation. In the unfixed condition, the incus–umbo ratio remains fairly constant up to 3000 Hz, meaning that the incus velocity is at a constant value below the umbo velocity. Above 3000 Hz the incus and umbo begin to behave differently. The ratio of their velocities moves towards zero and, in some temporal bones, becomes positive as the incus velocity exceeds that of the umbo. For three of the temporal bones the incus–umbo ratio in the unfixed state was –4 to –6 dB in the lower frequencies, while for the other it was around –12 dB.
Magnetically driven middle ear ossicles
355
10
Velocity (mm/s)
1
TB1 TB2 TB3 TB4
Unfixed Fixed
0.1 0.01 0.001
1000
500
2000
4000
8000
Frequency (Hz) Figure 3. Incus long process velocity before and after artificial fixation. Darker lines indicate the unfixed state and lighter lines the fixed state. The change in velocity of the incus is much greater than that observed for the umbo measurements.
After partial incus fixation, the ratio in the low and high frequency regions falls well below the unfixed response for all temporal bones. The ratio now lies between –20 and –40 dB. At this lowest value the incus long process velocity has sunk at some places to the level of the background noise, which was established by measuring on the promontory, away from the ossicles (not shown). At the very low frequencies (below 700 Hz), the noise levels begin to increase, and the data are of little interest. Discussion We first consider the practicality of the method. The set-up can easily be implemented in a clinical setting as the vibrometer can be combined with a surgical microscope without interfering with other procedures. The magnet, coil, and reflective tape can easily be sterilized. The cement used to attach the magnet is also physiologically harmless; it is commonly used as a restorative material in dentistry. It is known to be biocompatible and has previously been used in ME surgery [6]. A number of techniques were assessed
before luting cement was chosen to fix the magnet. Luting cement was found to provide good attachment of the magnet to bone and could easily be removed with forceps. It was found that the reflective tape could be placed on the ossicles and recovered without problems. Therefore, no aspect of the technique prevents its clinical application. The actual time required for the complete measurements (acoustic, calibration, and magnetic measurement) can be less than 1 min. With the additional time needed to place the reflective tape, attach the magnet, and position the apparatus, etc., the total time for the procedure should add no more than 10 min to the operation time. The major part of this lies in fixing the magnet, the cement normally taking about 5 min to set. Each measurement took approximately 6 s. The results presented here show the average for three repeated measurements. Additional measurements help to eliminate some noise, but a single measurement is generally sufficient to acquire useable data. The average variation from the mean for the three measurements was less than 5 dB, the largest part of
Incus – umbo ratio (dB)
10 0 –10 –20 –30 –40 500
TB1 TB2 TB3 TB4
Unfixed Fixed
1000
2000
4000
8000
Frequency (Hz) Figure 4. Incus–umbo ratio for the four temporal bones (TBs) before and after fixation. Darker lines indicate the unfixed state and lighter lines the fixed state. A clear drop in the ratio is observed in all TBs for all frequencies.
356
J. Peacock et al.
which comes from noise in the lower frequencies (below 1000 Hz). In the frequency range between 1 and 4 kHz the average variation from the mean was less than 1 dB. The method only allows measurement of ossicular vibration along one axis, parallel to the ear canal. This is probably the reason why, in the high frequencies, the incus velocity exceeds that of the umbo. The method is not intended to describe the true threedimensional motion of the ossicles, the aim of this work is to investigate if, within the constraints of measuring through the ear canal, it is possible to discriminate between a normal ear and one with fixed ossicles. Being restricted to measuring through the ear canal means that the range of angles for LDV access is limited, and measurements of different individuals should be comparable. Measurements were made on four fresh-frozen human temporal bones, the behavior of which is expected to be largely the same as for fully fresh material [7]. The data in Figure 2 show the umbo velocity in two of the temporal bones (TB2 and TB3) decreasing dramatically after partial incus fixation, but in the others the response does not deviate significantly from the unfixed situation. The difference is likely due to the precise nature of the fixation we made, in TB1 and TB4 the fixation was likely less complete than in the others. It would be difficult to determine how, or if, the state of ossicular mobility was affected by these data. Therefore, simple umbo response measurements are of limited clinical relevance in this case. Incus measurements are required to determine the fixation clearly. The decrease in incus motion after partial fixation is clear from Figure 3. The velocity at the tip of the incus long process decreases by a significant amount below the level of the unfixed condition. However, measuring incus velocity is not possible without opening the ME and so an easy method to achieve this that allows the ossicles to be stimulated in the same way as with an acoustic signal is required. The method of magnetic stimulation is one way to do this. Measuring the umbo response under acoustic stimulation, and then adapting the magnetic driving signal in such a way that umbo vibrations are the same after the TM has been elevated, ensures that the patient’s cochlea is not subject to mechanical overload during the measurement. Using the magnet technique allows us to measure the ratio of incus to umbo long process velocity, which is shown in Figure 4. Below 3000–4000 Hz, the ratio in the unfixed state is fairly constant and does not differ much between the temporal bones. The exception was TB1, for which the ratio was about 6 dB lower than in the
others. This may mean that the ear was naturally affected by a mild fixation. As the temporal bones were obtained through a tissue bank, audiometric data for the donors are, obviously, not available. Whatever the case, after partial fixation the ratio decreased for all temporal bones, reaching as low as –30 to –40 dB. This represents a clear difference of up to 30 dB between the partially fixed and unfixed conditions. The clearest data were obtained in the frequency region between 1000 and 4000 Hz. Below 700 Hz the data became increasingly noisy and above 4000 Hz the behavior of the incus velocity, relative to the umbo, changes with frequency and more variation is seen between different temporal bones. The measurements demonstrate that luting cement can be used to model partial incus fixation. However, it was not possible to make fixations with controlled steps to simulate well-defined degrees of fixation. It is likely that the smaller changes observed in TB1 and TB4 are due to a lower degree of fixation being achieved, as compared with the other two temporal bones. The procedure used to apply the cement will need to be further optimized in order to model different degrees of fixation in a repeatable way. Although the technique has only been demonstrated for one type of fixation, many others exist and the technique may also prove useful in assessing these conditions. However, much further work will be required before this can be fully established. In the clinical situation, a surgeon will not be able to see and compare data before and after the fixation. Therefore it is important for this technique that a base condition can be defined for measurements between different points on the ossicles. The natural variation in ossicular velocity between individuals (which can be 10 dB or more) makes defining a base velocity impossible, and limits the use of simple velocity measurements in detecting partial fixations. However, measuring the ratio between points may well provide an opportunity to define a base that is the same for all. The measurements presented in this paper hinted at the above, with the ratio of incus to umbo velocity being around 7 dB in the middle frequencies and also showing some similarity in the higher frequencies. However, much further work will be necessary to establish the base condition for the incus–umbo ratio and for the ratio between other points, and to establish what effects different types of partial ossicular fixation (different degrees and different locations) have on this ratio. Summary and conclusions Magnetic stimulation has proven an effective means to replicate the acoustic vibration response of the umbo after the TM has been elevated and allows
Magnetically driven middle ear ossicles for measurements of the incus through the ear canal. The technique is reliable, and easy to implement and perform. Measurements on fresh frozen human temporal bones show that for vibration measurements on the umbo, inter-subject variability can be larger than the changes induced by partial fixation of the incus. The fixation-produced changes are, however, clearly visible from measurements of the ratio of incus and umbo vibration level, with a decrease in the order of 10– 30 dB being seen. From the results presented in this paper, a ratio on the order of –4 to –12 dB between 1000 and 4000 Hz seems to be normal, and values outside this can be considered as abnormal. In this first investigation we have focused on fixations in the attic, which is readily accessible for ossicular fixation in temporal bones. More work is needed to measure the effects of artificially fixing different locations and for different degrees of fixation. Investigations with fixation of the stapes and of the anterior mallear ligament are under way.
Acknowledgments The authors wish to thank Fred Wiese, William Deblauwe, and Jan van Hecke for their assistance in building the measurement set-up. Thanks also to Pia Hjert, PO Medica, Sweden, for logistic assistance. This work was supported by the Research Foundation
357
Flanders - Fonds Wetenschappelijk Onderzoek (FWO) and funds from Akershus University Hospital. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References [1] Huber A, Linder T, Ferrazzini M, Schmid S, Dillier N, Stoeckli S, et al. Intraoperative assessment of stapes movement. Ann Otol Rhinol Laryngol 2001;110:31–5. [2] Chien W, Rosowski JJ, Ravicz ME, Rauch SD, Smullen J, Merchant SN. Measurements of stapes velocity in live human ears. Hear Res 2009;249:54–61. [3] Huber A, Schwab C, Linder T, Stoeckli SJ, Ferrazzini M, Dillier N, et al. Evaluation of eardrum laser doppler interferometry as a diagnostic tool. Laryngoscope 2001;111: 501–7. [4] Rosowski JJ, Nakajima HH, Merchant SN. Clinical utility of laser-Doppler vibrometer measurements in live normal and pathologic human ears. Ear Hear 2008;29:3–19. [5] Koike T, Murakoshi M, Hamanishi S, Yuasa Y, Yuasa R, Kobayashi T, et al. An apparatus for diagnosis of ossicular chain mobility in humans. Int J Audiol 2006;45:121–8. [6] Brask T. Reconstruction of the ossicular chain in the middle ear with glass ionomer cement. Laryngoscope 1999; 109:573–6. [7] Rosowski JJ, Davis PJ, Merchant SN, Donahue KM, Coltrera MD. Cadaver middle ears as models for living ears: comparisons of middle-ear input immittance. Ann Otol Rhinol Laryngol 1990;99:403–12.
Copyright of Acta Oto-Laryngologica is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
ORIGINAL ARTICLE
Magnetically driven middle ear ossicles with laser vibrometry as a new diagnostic tool to quantify ossicular fixation JOHN PEACOCK1, JORIS DIRCKX1 & MAGNUS VON UNGE2 1
Laboratory of Biomedical Physics, University of Antwerp, Belgium and 2Department of Otorhinolaryngology, Akershus University Hospital and University of Oslo, Oslo, Norway
Abstract Conclusion: Information on the degree of incus fixation can be gathered by measuring the ratio of incus to umbo long process velocity through the ear canal. Objectives: To test a new method of quantifying partial ossicular fixation in an ear with an elevated tympanic membrane. Methods: Measurements were made on four fresh-frozen human temporal bones. After elevating the tympanic membrane a small magnet was attached to the manubrium and an electromagnetic excitation coil was used to vibrate the ossicles. The vibration response of the tip of the incus long process and the umbo were measured before and after artificially fixating the incus to the lateral attic wall. Results: Partial incus fixation resulted in a decrease in both the incus and umbo velocities, with the incus velocity being more severely reduced. The decreased ratio of their vibrations is a clear indicator of the degree of incus fixation.
Keywords: Otitis media, ear surgery, incus fixation, umbo long process
Introduction A variety of conditions can result in a stiffening of the middle ear (ME) ossicles, which leads to hearing impairment. The ossicles may be more or less fixated due to various kinds of pathologies such as otosclerosis and malformations, and sequels of chronic otitis media such as tympanosclerosis and adhesions. In order to remedy a patient’s hearing loss surgical intervention is often required, and in order to determine the best course of surgical treatment some knowledge of the degree of ossicular mobility is useful. The final ossicular assessment is very often made by means of manual palpation during surgical exploration. Such assessment is difficult and the exact state of the ossicles can still remain unclear. Furthermore, manual pressing tests the ME’s quasi-static mechanical properties, which are not necessarily the same as the dynamic properties that govern sound transmission. A more objective method to assess mobility would be helpful.
A laser Doppler vibrometer (LDV) is a device that allows high precision noncontact vibration measurements on the nanometre scale. It has become a standard tool to measure ossicular motion, having been used to study ME mechanics in animals, human cadaveric temporal bones, and even in live human subjects [1–4]. Yet, despite the evident value of the technique, it has not found its place in routine clinical practice. Vibration measurements at different points along the ossicular chain can be made in temporal bones after drilling an opening in the facial recess. This method has been used to make measurements of stapes motion in live human subjects undergoing cochlear implantation [1,2]. However, in most operations this approach is too invasive, and for diagnostic purposes a means of measuring partial ossicular fixation through the ear canal would be preferable. During endaural surgery, the tympanic membrane (TM) is elevated giving free visual access to a large portion of the ossicular chain. However, in contrast to
Correspondence: Mr John Peacock, Laboratory of Biomedical Physics, University of Antwerp, Groenenborgerlaan 171, B-2020, Antwerp, Belgium. Tel: +32 (0) 3 265 34 38. E-mail: [email protected]
(Received 6 June 2013; accepted 23 August 2013) ISSN 0001-6489 print/ISSN 1651-2251 online 2014 Informa Healthcare DOI: 10.3109/00016489.2013.841990
Magnetically driven middle ear ossicles the situation above, where the TM can be driven by acoustic stimulation from the ear canal, the endaural surgery with elevation of the tympano-meatal flap prevents this approach. Some authors have indicated that measurements on the TM itself (before surgical exploration) can be used to evaluate the degree of ossicular fixation [3,4], but in many cases success is limited as partial ossicular fixation does not always result in reduced umbo velocities that are large enough to discriminate stiffened ossicles from normal ones, or to discriminate different types of pathology. Others have developed devices to measure partial fixation after elevating the TM [5], but this involves making physical contact with an actuator. The method has not gained widespread acceptance, possibly because special skill is required to make reliable mechanical contact between the actuator and the ossicles without overloading or hampering their motion. This study describes a new technique to measure ossicular vibration. The technique makes use of a small magnet and coil to vibrate the umbo after the TM is elevated and dissected free from the handle of the malleus, thus allowing for measurements of incus vibration through the ear canal. In this way data may be obtained in live humans with limited surgical intervention. The potential clinical usefulness of the technique is discussed. Material and methods Measurement set-up Four fresh frozen human temporal bones, acquired from Medcure Inc., Portland, Oregon, USA, were used. Before beginning any measurements they were left at room temperature for several hours to defrost, and were examined under an operating microscope for signs of abnormality. The measurement set-up is illustrated in Figure 1. A measurement set-up was built that consisted of a laser Doppler vibrometer (OFV-534, Polytec GmbH, Waldbronn, Germany) coupled to a surgical microscope (OPMI Sensera/S7; Carl Zeiss, Oberkochen, Germany). A small mirror, the orientation of which was controlled by piezo motors, was placed in front of the vibrometer, and could be used to accurately position the laser beam (spot size 200 mm at a distance of 30 cm). For comparison to stimulation with a magnet, an acoustic stimulus was provided by a free-field loudspeaker system, the output of which was measured by a probe microphone (Bruel & Kjær probe microphone, type 4182; Nærum, Denmark). The signal was designed in a PC using MATLAB and generated
353
Surgical microscope Vibrometer Temporal bone Speaker
Probe microphone
Coil
Figure 1. Drawing of the set-up. The dotted line shows the path of the laser beam.
using an A/D-D/A conversion board (RME HDSP 9632; Audio AG, Hainhausen, Germany) with 24-bit resolution and a sampling rate of up to 192 kHz. A stepped sine signal consisting of logarithmically spaced frequencies (with 16 per octave) between 250 and 8000 Hz was used. The sound pressure level (SPL) was adapted at each frequency to match an equal loudness curve at 80 phons (the relationship between loudness and SPL for humans is defined in ISO norm 226:2003). With an SPL equal for all frequencies, not all frequencies will sound equally loud to the human ear. By adapting the signal to the ear’s characteristics, it is possible to use higher stimulation levels at high and low frequencies without causing damage. An equal loudness level of 80 phons means that tones at all frequencies will sound, to the human ear, equally as loud as a 1000 Hz tone at 80 dB SPL. A small custom-made neodymium magnet (1 mm diameter 0.5 mm thickness, 0.003 g), in combination with a custom-built pancake coil (100 mm inner and 200 mm outer diameter, 97 spiral coiled windings) was used for the magnetic stimulation. The coil was driven by a power amplifier (DP P-900 Vintage Stereo Amplifier; Highlite International B.V., Kerkrade, the Netherlands) with the signal being constructed in MATLAB, the same as for the acoustic signal.
Measurement procedure To expose the ossicles and allow artificial fixation, the attic of the ME cleft was opened by drilling a hole in the tegmen tympani from the middle cranial fossa side. The approximate size of the hole was 5 mm. The
354
J. Peacock et al. 10
Velocity (mm/s)
1
TB1 TB2 TB3 TB4
Unfixed Fixed
0.1 0.01 0.001
500
1000
2000
4000
8000
Frequency (Hz) Figure 2. Umbo velocity before and after artificial fixation. The darker lines indicate the unfixed state and the lighter lines the fixed state. In temporal bone (TB) 1 and TB4 the velocity in the fixed state is in the same range as the unfixed condition for the other two bones.
hole gave access to the short process of the incus, the incus body, and the head of the malleus. Sound was presented to the TM, and the sound levels were calibrated using the probe microphone with its tip placed close to the opening of the ear canal. The calibrated sound was then used to stimulate the TM and data on the vibration response of the umbo were measured by the vibrometer. To improve the strength of the reflected vibrometer signal, a small piece of reflective tape (Polytec) was placed at the umbo. The TM was then elevated and pieces of reflective tape were placed on the umbo and at the tip of the long process of the incus. The magnet was then attached to the manubrium, at a point close to the umbo, using a small drop of glass ionomer luting cement (GC Corporation, Tokyo, Japan). It takes a few minutes to set and fixes the magnet quite rigidly. The custom-built pancake coil was placed over the ear, with the opening to the ear canal near its center, as if the coil was resting on a patient’s head. A signal was then put to the coil to vibrate the magnet, and the vibrometer was again used to measure the response of the umbo. The vibration response measured with magnetic stimulation was compared to that achieved with the earlier acoustic stimulation and the signal to the coil was adjusted, by means of an iterative procedure, until the umbo response matched that achieved acoustically to within 1 dB. Three separate measurements were made at each measurement point (three at the umbo and three at the tip of the long process of the incus). Subsequently, the incus was artificially fixed by the application of luting cement (the same kind as was used to attach the magnet) between its head and the wall of the attic bone. A drop of cement was applied to fill out the space between its most superior portions (the head and short process) and the lateral attic wall of the temporal bone. After the cement had set the
measurements were repeated as before. The time taken for each measurement was just under 6 s. Results All four temporal bones were examined under an operating microscope. The TM was examined and appeared normal, and after TM elevation the ME showed no signs of abnormality. The mucosa was examined for adhesions or scar tissues and the ossicles appeared normal. No other signs of pathology were detected. Figure 2 shows the velocity of the umbo before and after partial incus fixation for all four temporal bones (labelled TB1–4). In TB2 and TB3 the vibration response clearly diminishes after the cement is applied. In TB1 and TB4 the response does drop, as compared with the original vibration level in each temporal bone, but the vibration level after fixation in these specimens does not differ significantly from the unfixed vibration level. Figure 3 shows the velocity of the incus long process before and after partial incus fixation for all four temporal bones. In contrast to measurements on the umbo, the response of all temporal bones clearly diminishes after applying the cement and the change in response is larger than for the umbo. Figure 4 shows the ratio of incus long process to umbo velocity before and after fixation. In the unfixed condition, the incus–umbo ratio remains fairly constant up to 3000 Hz, meaning that the incus velocity is at a constant value below the umbo velocity. Above 3000 Hz the incus and umbo begin to behave differently. The ratio of their velocities moves towards zero and, in some temporal bones, becomes positive as the incus velocity exceeds that of the umbo. For three of the temporal bones the incus–umbo ratio in the unfixed state was –4 to –6 dB in the lower frequencies, while for the other it was around –12 dB.
Magnetically driven middle ear ossicles
355
10
Velocity (mm/s)
1
TB1 TB2 TB3 TB4
Unfixed Fixed
0.1 0.01 0.001
1000
500
2000
4000
8000
Frequency (Hz) Figure 3. Incus long process velocity before and after artificial fixation. Darker lines indicate the unfixed state and lighter lines the fixed state. The change in velocity of the incus is much greater than that observed for the umbo measurements.
After partial incus fixation, the ratio in the low and high frequency regions falls well below the unfixed response for all temporal bones. The ratio now lies between –20 and –40 dB. At this lowest value the incus long process velocity has sunk at some places to the level of the background noise, which was established by measuring on the promontory, away from the ossicles (not shown). At the very low frequencies (below 700 Hz), the noise levels begin to increase, and the data are of little interest. Discussion We first consider the practicality of the method. The set-up can easily be implemented in a clinical setting as the vibrometer can be combined with a surgical microscope without interfering with other procedures. The magnet, coil, and reflective tape can easily be sterilized. The cement used to attach the magnet is also physiologically harmless; it is commonly used as a restorative material in dentistry. It is known to be biocompatible and has previously been used in ME surgery [6]. A number of techniques were assessed
before luting cement was chosen to fix the magnet. Luting cement was found to provide good attachment of the magnet to bone and could easily be removed with forceps. It was found that the reflective tape could be placed on the ossicles and recovered without problems. Therefore, no aspect of the technique prevents its clinical application. The actual time required for the complete measurements (acoustic, calibration, and magnetic measurement) can be less than 1 min. With the additional time needed to place the reflective tape, attach the magnet, and position the apparatus, etc., the total time for the procedure should add no more than 10 min to the operation time. The major part of this lies in fixing the magnet, the cement normally taking about 5 min to set. Each measurement took approximately 6 s. The results presented here show the average for three repeated measurements. Additional measurements help to eliminate some noise, but a single measurement is generally sufficient to acquire useable data. The average variation from the mean for the three measurements was less than 5 dB, the largest part of
Incus – umbo ratio (dB)
10 0 –10 –20 –30 –40 500
TB1 TB2 TB3 TB4
Unfixed Fixed
1000
2000
4000
8000
Frequency (Hz) Figure 4. Incus–umbo ratio for the four temporal bones (TBs) before and after fixation. Darker lines indicate the unfixed state and lighter lines the fixed state. A clear drop in the ratio is observed in all TBs for all frequencies.
356
J. Peacock et al.
which comes from noise in the lower frequencies (below 1000 Hz). In the frequency range between 1 and 4 kHz the average variation from the mean was less than 1 dB. The method only allows measurement of ossicular vibration along one axis, parallel to the ear canal. This is probably the reason why, in the high frequencies, the incus velocity exceeds that of the umbo. The method is not intended to describe the true threedimensional motion of the ossicles, the aim of this work is to investigate if, within the constraints of measuring through the ear canal, it is possible to discriminate between a normal ear and one with fixed ossicles. Being restricted to measuring through the ear canal means that the range of angles for LDV access is limited, and measurements of different individuals should be comparable. Measurements were made on four fresh-frozen human temporal bones, the behavior of which is expected to be largely the same as for fully fresh material [7]. The data in Figure 2 show the umbo velocity in two of the temporal bones (TB2 and TB3) decreasing dramatically after partial incus fixation, but in the others the response does not deviate significantly from the unfixed situation. The difference is likely due to the precise nature of the fixation we made, in TB1 and TB4 the fixation was likely less complete than in the others. It would be difficult to determine how, or if, the state of ossicular mobility was affected by these data. Therefore, simple umbo response measurements are of limited clinical relevance in this case. Incus measurements are required to determine the fixation clearly. The decrease in incus motion after partial fixation is clear from Figure 3. The velocity at the tip of the incus long process decreases by a significant amount below the level of the unfixed condition. However, measuring incus velocity is not possible without opening the ME and so an easy method to achieve this that allows the ossicles to be stimulated in the same way as with an acoustic signal is required. The method of magnetic stimulation is one way to do this. Measuring the umbo response under acoustic stimulation, and then adapting the magnetic driving signal in such a way that umbo vibrations are the same after the TM has been elevated, ensures that the patient’s cochlea is not subject to mechanical overload during the measurement. Using the magnet technique allows us to measure the ratio of incus to umbo long process velocity, which is shown in Figure 4. Below 3000–4000 Hz, the ratio in the unfixed state is fairly constant and does not differ much between the temporal bones. The exception was TB1, for which the ratio was about 6 dB lower than in the
others. This may mean that the ear was naturally affected by a mild fixation. As the temporal bones were obtained through a tissue bank, audiometric data for the donors are, obviously, not available. Whatever the case, after partial fixation the ratio decreased for all temporal bones, reaching as low as –30 to –40 dB. This represents a clear difference of up to 30 dB between the partially fixed and unfixed conditions. The clearest data were obtained in the frequency region between 1000 and 4000 Hz. Below 700 Hz the data became increasingly noisy and above 4000 Hz the behavior of the incus velocity, relative to the umbo, changes with frequency and more variation is seen between different temporal bones. The measurements demonstrate that luting cement can be used to model partial incus fixation. However, it was not possible to make fixations with controlled steps to simulate well-defined degrees of fixation. It is likely that the smaller changes observed in TB1 and TB4 are due to a lower degree of fixation being achieved, as compared with the other two temporal bones. The procedure used to apply the cement will need to be further optimized in order to model different degrees of fixation in a repeatable way. Although the technique has only been demonstrated for one type of fixation, many others exist and the technique may also prove useful in assessing these conditions. However, much further work will be required before this can be fully established. In the clinical situation, a surgeon will not be able to see and compare data before and after the fixation. Therefore it is important for this technique that a base condition can be defined for measurements between different points on the ossicles. The natural variation in ossicular velocity between individuals (which can be 10 dB or more) makes defining a base velocity impossible, and limits the use of simple velocity measurements in detecting partial fixations. However, measuring the ratio between points may well provide an opportunity to define a base that is the same for all. The measurements presented in this paper hinted at the above, with the ratio of incus to umbo velocity being around 7 dB in the middle frequencies and also showing some similarity in the higher frequencies. However, much further work will be necessary to establish the base condition for the incus–umbo ratio and for the ratio between other points, and to establish what effects different types of partial ossicular fixation (different degrees and different locations) have on this ratio. Summary and conclusions Magnetic stimulation has proven an effective means to replicate the acoustic vibration response of the umbo after the TM has been elevated and allows
Magnetically driven middle ear ossicles for measurements of the incus through the ear canal. The technique is reliable, and easy to implement and perform. Measurements on fresh frozen human temporal bones show that for vibration measurements on the umbo, inter-subject variability can be larger than the changes induced by partial fixation of the incus. The fixation-produced changes are, however, clearly visible from measurements of the ratio of incus and umbo vibration level, with a decrease in the order of 10– 30 dB being seen. From the results presented in this paper, a ratio on the order of –4 to –12 dB between 1000 and 4000 Hz seems to be normal, and values outside this can be considered as abnormal. In this first investigation we have focused on fixations in the attic, which is readily accessible for ossicular fixation in temporal bones. More work is needed to measure the effects of artificially fixing different locations and for different degrees of fixation. Investigations with fixation of the stapes and of the anterior mallear ligament are under way.
Acknowledgments The authors wish to thank Fred Wiese, William Deblauwe, and Jan van Hecke for their assistance in building the measurement set-up. Thanks also to Pia Hjert, PO Medica, Sweden, for logistic assistance. This work was supported by the Research Foundation
357
Flanders - Fonds Wetenschappelijk Onderzoek (FWO) and funds from Akershus University Hospital. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References [1] Huber A, Linder T, Ferrazzini M, Schmid S, Dillier N, Stoeckli S, et al. Intraoperative assessment of stapes movement. Ann Otol Rhinol Laryngol 2001;110:31–5. [2] Chien W, Rosowski JJ, Ravicz ME, Rauch SD, Smullen J, Merchant SN. Measurements of stapes velocity in live human ears. Hear Res 2009;249:54–61. [3] Huber A, Schwab C, Linder T, Stoeckli SJ, Ferrazzini M, Dillier N, et al. Evaluation of eardrum laser doppler interferometry as a diagnostic tool. Laryngoscope 2001;111: 501–7. [4] Rosowski JJ, Nakajima HH, Merchant SN. Clinical utility of laser-Doppler vibrometer measurements in live normal and pathologic human ears. Ear Hear 2008;29:3–19. [5] Koike T, Murakoshi M, Hamanishi S, Yuasa Y, Yuasa R, Kobayashi T, et al. An apparatus for diagnosis of ossicular chain mobility in humans. Int J Audiol 2006;45:121–8. [6] Brask T. Reconstruction of the ossicular chain in the middle ear with glass ionomer cement. Laryngoscope 1999; 109:573–6. [7] Rosowski JJ, Davis PJ, Merchant SN, Donahue KM, Coltrera MD. Cadaver middle ears as models for living ears: comparisons of middle-ear input immittance. Ann Otol Rhinol Laryngol 1990;99:403–12.
Copyright of Acta Oto-Laryngologica is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
