Percutaneous vs, transcutaneous transducers for hearing by direct bone conduction 6 0 HAUNSSON, PhD, ANDERS TJELLSTROM, MD and PEDER CARLSSON, MSc, Goteborg, Sweden
There is a substantial need for improvement of the hearing situation for patients with chronic middle ear or ear canal disorders. To improve hearing for these patients, two different bone conduction hearing systems have been developed. The Nobelpharma Auditory System HC 200-the bone-anchored hearing aid we present here-uses a percutaneoustransducer; whereas the Audiant device, developed by Dr. Jack Hough, uses a transcutaneous transducer. In percutaneous transmission, the transducer is directly coupled to the bone by means of a permanent skin penetration, whereas in transcutaneoustransmission one part of the transducer is implanted and the other part Is kept outside the intact skin and soft tissue. Comprehensive audiologic assessments indicate great differences in performance between the two systems. These differences probably originate in differencesin length of gap and in different suspension properties of the two transducer systems. This article will demonstrate that large gaps, such as in the transcutaneoustransducer, can be devastating for power consumption, maximum output capability, and second harmonic distortion. Since the properties of the suspension in the transcutaneous transducer are not under adequate control and the complication risk of permanent skin penetration is low, we continue to concentrate our efforts on percutaneous transducer systems. (OTOLARYNGOL HEAD NECKSURG 1990;102:339,]
I n spite of the many recent achievements in reconstructive middle ear surgery, there is still a need for amplification in selected cases. Conventional bone conduction hearing aids have been in poor reputation for many years. The externally worn transducer is continuously pressed against the intact skin and soft tissues either by a steel spring or by the arms of a pair of glasses. The static pressure developed by the spring should exceed two newtons. Discomfort, poor sound quality, and cumbersomeness have been some of the drawbacks. A totally or partly implantable hearing aid for patients with chronic middle ear canal disorders has been a goal for many years, but is still unavailable. In 1977, we started to fit such patients with a percutaneous direct bone conduction hearing system-the Nobelpharma Auditory System HC 200 (Nobelpharma AB, Goteborg, Sweden). This system consists of two parts; namely, an external part-the HC 200 sound processor (also ~
~
From the Department of Applied Electronics, Chatmers University of Technology (Drs. Hikansson and Carlsson), and the ENT Department, Sahlgren Hospital (Dr.Tjellstrom). Submitted for publication R b . 16, 1989; accepted Sept. 18, 1989. Reprint requests: Bo Hikansson, PhD, Department of Applied Electronics, Chalmers University of Technology, S-412 96 Goteborg, Sweden.
23/1/1m
known as the bone-anchored hearing aid)-which is coupled to an implanted part, the Brinemark System. The implanted part consists of a titanium fixture placed and integrated in the mastoid process, and a skinpenetrating titanium abutment.' Since 1977, more than 300 patients in Sweden and 200 patients in other countries have been provided with this type of hearing system, and numerous reports have been published presenting the surgical technique,* skin reaction^,^.^ and audiologic data.'-I0 Our experience with a percutaneous coupling to the bone is that in 92% of the cases," patients with a sensorineural hearing loss down to 45 dB PTA (pure tone average 0.5, 1, 2, 3 kHz) could benefit from the HC 200 sound processor when compared with their old hearing devices. Moreover, speech discrimination tests performed with background noise show that the HC 200 sound processor has a significantly better noise endurance for equal intelligibility than conventional bone conduction hearing aids. The credit for making hearing through bone conduction accepted in North America belongs to Dr. Jack Hough, MD, of Oklahoma City, and his co-workers. This group has developed a transcutaneous bone conductor, the Audiant device (Xomed Inc., Jacksonville, Ha.), which allows the skin and soft tissues to remain intact. The subcutaneous part of this device is a rareearth magnet, housed in a titanium disk, sealed and 339
340
HkANSSON et al.
OtolaryngologyHead and Neck Surgery
aim of this article is to theoretically-as well as experimentally-describe and discuss what effects these differences between the two systems can have on sound transmission quality and power requirements. TECHNICAL BACKGROUND Transducers for Hearing Aids
Fig. I.Schematic drawing illustrates components and parameters of fundamental importance in the performance of a variable reluctance transducer.
potted in medical grade silicone, and attached to the temporal bone by a Herbert orthopedic screw. Such a device was first implanted in the mastoid process above the sinodural angle in a patient in 1984. In 1986, the first paper was published describing the technique and the hearing results from the first 10 patients.I2 Among the audiologic criteria for implantation were bone conduction thresholds no greater than 30 dB PTA (0.5, 1, 2 kHz) and at no single frequency greater than 40 dB. In the discussion of that paper, the authors state that one of the major problems is associated with the power needed. At the American Academy meeting in Chicago in 1987, Dr. Houghi3stated (in his instructional session) that the bone conduction thresholds should not exceed 25 dB PTA and that the patients had difficulties excluding background sound from the subject of interest. Since 1984, approximately 500 patients in the United States and 100 patients in other countries have been provided with this type of bone conductor.i4 Although the basic principles of the transducers are the same (variable reluctance) in the two hearing devices, there are great differences in performance, as indicated earlier. The differences in performance between a percutaneous and a transcutaneous transducer are probably explained by the great difference in length of gap in the magnetic circuit, and the difference in suspension properties. The length of gap and the suspension properties are of fundamental importance for the performance of a variable reluctance transducer. The
Output transducers for hearing aids are named differently, depending on their applications. In airconduction hearing aids, receiver or earphone are the most appropriate names, whereas in bone-conduction hearing aids, names such as vibrator, bone conductor, or transducer are commonly used. For our purposes, the name transducer is used, and the discussion is restricted to transducers for bone conduction hearing aids. In the design of electromechanical transducers for bone conduction hearing aids, a solution that combines properties such as small size, wide frequency range, high output capability, low level of distortion, and low consumption of current is of utmost importance. Piezoelectric and magnetostrictive transducers are generally considered inconvenient, basically because of poor low-frequency response. Electrodynamic transducers of the moving coil type, which are most common in ordinary loudspeakers, are not considered becausefor physical reasons-they have too low an electrical input impedance, which means they require relatively high current under ordinary working conditions. Ordinary hearing aid batteries are not designed for such working conditions and, moreover, high consumption of current causes severe voltage fluctuations that, as a secondary effect, cause internal electrical stability problems. But an electrodynamic transducer of the variable reluctance type combines properties such as small size, wide frequency range, low levels of current consumption, and excellent electroacoustic performance, which are desirable in hearing aid applications. Prlncipal Design of a Variable Reluctance Transducer Principle of operation. In the following material, some equations are presented showing the influence of length of gap and of suspension properties on the performance of a variable reluctance transducer. These equations are described in greater detail by Carlsson and Hiikansson.l5 As indicated earlier, electromechanical transducers of the variable reluctance typeschematically illustrated in Fig. 1-are commonly used in conventional bone conductors. The mass M, including the coil, forms the counterweight. This mass should include a permanent magnet that produces the static magnetic flux $o. It should be noted that for a class A output amplifier, normally used in hearing aids
Volume 102 Number 4 April 1990
Percutaneousvs. transcutaneous transducers 341
with low-power capability, the static magnetic flux could be wholly or partially produced by the electrical current drain. The total magnetic flux is the sum of the static flux +o and the signal flux produced by the electrical current i. It can be shown that the total force F between the two sides of the gap is proportional to (denoted a ) the square of the total magnetic flux, yielding the relation
+ +-
F (Y
+ 2cu,+- + + -’= F, + F- + Fd (equation I )
under the assumption that the leakage of magnetic flux is negligible. From equation 1, it is apparent that the total force is composed of three forces: the static force Fo, the signal force F-, and the second harmonic distortion force Fd. Moreover, it is clear from equation 1 that it is important to have a static flux that is high in comparison with the signal flux, in order to achieve a low relative second harmonic distortion (Fd/F- a +-/+o).
Influence of length of gap on electromechanical conversion. The gap 6, as defined in Fig. 1, is the length of the magnetic path consisting of nonmagnetic material. Normally this gap is air, but it could also consist of some combination of biologic and inorganic material of low permeability, such as brass or various permanent materials. The magnetic flux +o is proportional to the strength of the permanent magnet (i.e., size and BH-product) and reciprocal to the gap 6 according to equation 2, if it is assumed that all magnetic reluctance is concentrated to the gap and that leakage is negligible. Equation 2 is stated as cb0 a (strength of magnet)/S (equation 2).
The electromechanical conversion from the electrical input current i or input voltage v to the mechanical signal output force F- is of the utmost importance and determines the maximum output capability, as well as the consumption of current for a given signal force level. The signal flux is proportional to the input current i, but is reciprocal to 6 as follows:
+-
+- a i / 6
(equation 3 ) .
Combining equations 1, 2, and 3, we find the following relation for the signal force as a function of input current and gap, assuming the strength of magnet to be unchanged: F_ a i1S2
(equation 4 ) .
From equation 4 it is clear that a large 6 can be devastating for the electromechanical conversion efficiency. As the average consumption of current determines the battery consumption, a large 6 results in high
TlQJ&@
Fig. 2. The percutaneous transducer system: Skull bone (I),soft tissue (2).titanium fixture (3), titanium abutment (4), bayonet coupling (5Jand percutaneous transducer (6).
battery costs. Moreover, it can be seen from equation 4 that an equal signal force for various gaps 6 is obtained if the current i is changed by the same amount squared (i a 6*). Combining equations 1, 2, and 3 yields that the second harmonic distortion is proportional to the gap squared: Fd a 6’ . F?
(equation 5).
Hence large gaps lead to high levels of second harmonic distortion that can be very annoying to the patients. In hearing aids, the maximum output capability is normally determined by the battery voltage. The input current i in equation 4 can be transformed to the input voltage v by the reciprocal of the electrical input impedance, according to Ohm’s law. Assuming that the electrical input impedance is determined by the inductance only (electrical resistance and hysteresis losses are disregarded), and recalling that the inductance is reciprocally proportional to the gap 6, then the signal force yields F- a v/6
(equation 6 ) .
For a given battery voltage, the maximum signal force can be considerably reduced because of a large gap 6. Influence of suspension properties on mechanical transmission. The signal force created in the gap is transmitted through the mechanical components of the transducer to the load. The suspension spring, characterized by its compliance C and inherent damping R,
OtolaryngologyHead and Neck Surgery
342 WkANSSON et al.
gap therefore requires a suspension spring with wellcontrolled properties.
Fig. 3. The transcutaneous transducer system: Skull bone (I), soft tissue of reduced thickness (21, internal part of transcutaneous transducer (57, and external part of transcutaneous transducer (4).
in combination with the transducer mass M, determines the basic resonance frequency and damping of the transducer. This particular frequency and this particular damping are of fundamental importance for the electromechanical performance of the transducer because they determine the frequency region, which is given an enhanced electromechanical sensitivity. In the following material, it is assumed that the mass M is the same in the two systems discussed. The simplified relation for the resonance frequency f is fa1 1 s
(equation 7 ) ,
if we assume that the load impedance is much greater than the output impedance of the transducer, which is the case in direct bone conduction, according to H%ansson et al.5 Furthermore, the mqimum output F, from the transducer is determined by the damping coefficient a according to the following equation: 1 F,a-a-
a
1 RvZ
(equation 8 ) .
Equations 7 and 8 show that the dynamic behavior of a variable reluctance transducer is strongly influenced by the suspension properties, compliance C and damping R. The suspension is also very important from the static point of view. The force created by static deflection of the suspension spring counteracts the attraction force from the permanent magnet and, at a certain length of gap, equilibrium is established. A narrow and stable
Percutaneous vs. Transcutaneous Transducers The percutaneous transducer. In Fig. 2, the principle of a percutaneous transducer is shown. A titanium fixture is permanently implanted into the mastoid bone of the skull and a permanent skin penetration is made with a titanium abutment. A bayonet coupling is used for attachment of the percutaneous transducer. The airfilled gap of the transducer is adjusted to an appropriate length, and the resonance frequency and damping (the region of enhanced electromechanical sensitivity) are chosen to fit the amplification need for the typical hearing loss among the patients this hearing device is aimed for by selection of appropriate mass M, compliance C , and damper R. The transcutaneous transducer. In Fig. 3, the principle of a transcutaneous transducer is shown. The implanted part of the transcutaneous transducer is housed in biocompatible material and secured to the mastoid bone of the skull. The permanent magnet (or magnets), preferably housed in both the external part and the internal part of the transcutaneous transducer, not only have the purpose of reducing second harmonic distortion, but should also keep the device safely in place over the intact skin and soft tissue. For this transducer principle, the gap is filled with the skin and soft tissues of reduced thickness, the properties of which then determine the length of the gap, as well as the resonance frequency and the damping. DISCUSSION
The purpose of this article has been to discuss some of the fundamental properties of a variable reluctance transducer with reference to percutaneous vs. transcutaneous bone conduction. Transcutaneous transmission means that there is no metallic contact between the internal and external parts of the transducer-instead, there is intact soft tissue in between. The thickness of the soft tissue determines the gap length, which-in this situation-is relatively large. With a percutaneous transmission system, on the other hand, the gap between the two parts of the transducer is small and kept constant by a metallic spring suspension. It is apparent that, with certain assumptions, large gaps can be devastating for power consumption and maximum output capability, as well as for second harmonic distortion. One aspect that can influence these results is that the strength of the magnet is assumed to be the same for the two systems discussed. It appears from equations 1 and 2 that a strong magnet can solve most of the problems associated with large gaps. This is misleading, however, because an increased amount of permanent magnet material will significantly reduce
Volume 102 Number 4 April 1990
Percutaneous vs. transcutaneous transducers 343
100
200
500
lk Frequency Hz
2k
5k
1Ok
Fig. 4. Output force levels at 50 d B SPL input (lower broken and solid lines) and at saturation (upper broken and solid lines), produced by t h e Audiant A.T.E. sound processor (solid lines) and the HC 200
sound processor (broken lines). the dynamic transmission efficiency, since high quality permanent magnet materials (such as rare earth cobalt) have poor dynamic properties. More details of these aspects were given by Carlsson and Hakanss~n.'~ Also, the strength of the permanent magnets is closely related to the size (or weight) of the magnet. A large size is not desirable for aesthetic reasons, and a heavy weight leads to problems keeping the transcutaneous transducer in place. One assumption in the analysis was that the magnetic leakage flux was negligible. This is true for narrow gaps, but for wide gaps a portion of the signal flux passes outside the area of the gap; this implies a reduced dynamic transmission efficiency. The length of the gap, as well as the properties of the soft tissue maintaining the gap in a transcutaneous transducer, will vary individually. The tight scar tissue layer of an adult who has been exposed to previous surgery might have mechanical characteristics different from the tissue of a child who has not been operated on before. There could also be changes over time in the individual patient. From the use of conventional bone conduction hearing aids, it is well-known that the patients often will have impressions of the soft tissues because of the pressure needed. Many patients using such conventional bone conduction aids cannot have the transducer in the same position all day and will move it around and even take it off during certain periods. The force needed for a transcutaneous device will probably be of the same magnitude, and some of the
same problems could be anticipated. In conclusion, one might expect that the transcutaneous device will give relatively high intersubject as well as intrasubject variation in electromechanical transmission in comparison with the percutaneous device. The percutaneous transmission system has, of course, the potential disadvantage of inflammation and infection at the site of the skin penetration. The experience with percutaneous implants in the mastoid, however, goes back to 1977, and the frequency of skin problems has been low.' More than 360 patients have received a total of about 580 implants in the mastoid process for a direct bone conduction instrument or for retention of auricular prostheses. These patients have been followed carefully and the frequency of adverse skin reactions has been registered. More than 3000 observations have been made, and in 90% of those no adverse reactions of the skin around the skin penetration were noted. About 4% of the observations were of a type that called for extra visits and treatment. Only one of 250 implants for direct bone conduction had to be removed as a result of soft tissue problems. Preliminary Experimental Comparison
There are differences between the percutaneous and the transcutaneous hearing systems in terms of how to perform objective (patient-independent)quality control of the complete hearing device. For the percutaneous device, a separate unit, the skull simulator TU-1000,
344
&ANSON
OtolaryngologyHead and Neck Surgery
et al.
described by Hakansson and Carlsson,I6 has been developed. The skull simulator is capable of measuring the output force level from the HC 200 sound processor, with a load impedance relevant for the impedance of the normal patient. This measurement procedure cannot be carried out with the transcutaneous device in a similar way because a vital part of that transducer is totally implanted. In order to perform comparative experiments between the Audiant A.T.E. sound processor and the HC 200 sound processor, however, an adaptor was developed so that an implantable orthopedic screw of the Audiant device could be rigidly attached to the skull simulator TU 1000. The internal and the external units were separated by application of a 1.5 mm-thickness skin transplant in between. Results from measurements of the output force level produced by the two devices are presented in Fig. 4. The measurements were performed with all external adjustable controls set to produce maximum output force level. From the saturation force levels (SFL) shown in Fig. 4 by the upper curves, it was found that the HC 200 sound processor could produce approximately 20 dB higher maximum force levels than the Audiant device. This result is in agreement with the great difference in maximum allowable bone thresholds (45 vs. 25 dB PTA between the two hearing systems). The sound pressure level for the measurement of the linear frequency response had to be lowered from 60 dB, which is normally used for hearing aids, down to 50 dB to achieve sufficient linearity of the Audiant device. From the output force levels recorded at 50 dB (OFL,,) and the SFL it seems that the maximum dynamic range is considerably smaller for the Audiant device than for the HC 200 sound processor. The difference in consumption of current was measured with pure tones at l .O, l .6, and 2.5 kHz, respectively, and at an equal output level for the two devices when working in their linear range. It was found that the Audiant device (as an average over the three frequencies) required 26 times higher current than the HC 200 sound processor for an equal output force level. CONCLUSION
There is basically one reason to choose the transcutaneous device in preference to the percutaneous one-namely, to avoid permanent skin penetration. On the other hand, the percutaneous device offers the opportunity to optimize the transducer for a stable and well-controlled performance, and, above all, to reach high efficiency in electromechanical transmission. In view of the low complication rates for the permanent skin penetration and fixture implant, and the superior electromechanical performance, a percutaneous trans-
ducer system is chosen in preference to a transcutaneous one for the bone-anchored hearing aid, also known as the Nobelpharma Auditory System HC 200. We wish to acknowledge the skillful technical assistance of research engineer Berndt Anderson in the accomplishment of the experimental session. REFERENCES
1. Tjellstrom A, Lindstrom J, Halltn 0, Albrektsson T, Brinemark P-I. Osseointegrated titanium implants in the temporal bone. A clinical study on bone-anchored hearing aids. Am J Otol 1981;2:304-10. 2. Tjellstrom A. Osseointegrated systems and their application in the head and neck. Adv. Otolaryngol Head Neck Surg 1989; 3:39-70. 3. Holgers K-M, Tjellstrom A, Erlandson B-E, Bjursten LM. Soft tissue reactions around percutaneous implants: a clinical study on soft tissue conditions around skin penetrating titanium implants for bone-anchored hearing aids. Am J Otol 1988;9:56-9. 4. Jacobson M, Tjellstrom A, Tomsen P, Albrektson T. Soft tissue infection around a skin penetrating osseointegrated implant: a case report. Scand J Plast Reconstr Surg 1987;21:225-8. 5. Hikansson B, Carlsson P, Tjellstrom A. The mechanical point impedance of the human head, with and without skin penetration. J Acoust Soc Am 1986;801065-75. 6. Hikansson B, Tjellstrom A, Rosenhall U. Acceleration levels and thresholds with direct bone conduction versus conventional bone conduction. Acta Otolaryngologica (Stockh) 1985;100:24052. 7. H&ansson B, Tjellstrom A, Rosenhall U. Hearing thresholds with direct bone conduction versus conventional bone conduction. Scand Audiol 1984;13:3-13. 8. Carlsson P, Hikansson B, Rosenhall U, Tjellstrom A. A speech reception threshold test in noise with the bone-anchored hearing aid a comparative study. OTOLAFXNGOL HEAD NECK SURC 1986;94:421-6. 9. Tjellstrom A, Jacobson M. A 10-year experience with maxillofacial bone-anchored implants. Presented at the Annual Meeting of the American Academy of Otolaryngology-Head and Neck Surgery, Washington, D.C., Sept. 25-29, 1988. 10. Tjellstriim A, Jacobson M, Norvell B, Albrektsson T. Patient attitudes to the bone-anchored hearing aid: Results of a questionnaire study. Scand Audiol 1988;18:119-23. 11. Hikansson B, Lidtn G, Tjellstrom A, Ringdahl A, Carlsson P, Erlandsson B-E. Ten years of experience of the Swedish boneanchored hearing system. Ann Otol Rhino1 Laryngol: submitted 1989. 12. Hough J, Vernon J, Dormer K, Johnson B, Himelick T. Experiences with implantable hearing devices and a presentation of a new device. Ann Otol Rhino1 Laryngol 1986;95:60-5. 13. Hough JVD, Dormer KJ. Implantable hearing devices: XOMED bone conductor. American Academy of Otolaryngology-Head and Neck Surgery, Course 3337, 1987. 14. Campos CT. A chronology of an implantable bone conductor hearing device. Hear Instrum 1988;39:36, 38, 78. 15. Carlsson P,Hikansson B. A new transducer for hearing by direct bone conduction. Technical Report No. 2:88. Research Laboratory of Medical Electronics, Chalmers University of Technology, Goteborg, Sweden, 1988. ISBN 91-7546-050-5. 16. Hikansson B, Carlsson P. Skull simulator for direct bone conduction hearing devices. Scand Audiol 1988;18:91-8.
There is a substantial need for improvement of the hearing situation for patients with chronic middle ear or ear canal disorders. To improve hearing for these patients, two different bone conduction hearing systems have been developed. The Nobelpharma Auditory System HC 200-the bone-anchored hearing aid we present here-uses a percutaneoustransducer; whereas the Audiant device, developed by Dr. Jack Hough, uses a transcutaneous transducer. In percutaneous transmission, the transducer is directly coupled to the bone by means of a permanent skin penetration, whereas in transcutaneoustransmission one part of the transducer is implanted and the other part Is kept outside the intact skin and soft tissue. Comprehensive audiologic assessments indicate great differences in performance between the two systems. These differences probably originate in differencesin length of gap and in different suspension properties of the two transducer systems. This article will demonstrate that large gaps, such as in the transcutaneoustransducer, can be devastating for power consumption, maximum output capability, and second harmonic distortion. Since the properties of the suspension in the transcutaneous transducer are not under adequate control and the complication risk of permanent skin penetration is low, we continue to concentrate our efforts on percutaneous transducer systems. (OTOLARYNGOL HEAD NECKSURG 1990;102:339,]
I n spite of the many recent achievements in reconstructive middle ear surgery, there is still a need for amplification in selected cases. Conventional bone conduction hearing aids have been in poor reputation for many years. The externally worn transducer is continuously pressed against the intact skin and soft tissues either by a steel spring or by the arms of a pair of glasses. The static pressure developed by the spring should exceed two newtons. Discomfort, poor sound quality, and cumbersomeness have been some of the drawbacks. A totally or partly implantable hearing aid for patients with chronic middle ear canal disorders has been a goal for many years, but is still unavailable. In 1977, we started to fit such patients with a percutaneous direct bone conduction hearing system-the Nobelpharma Auditory System HC 200 (Nobelpharma AB, Goteborg, Sweden). This system consists of two parts; namely, an external part-the HC 200 sound processor (also ~
~
From the Department of Applied Electronics, Chatmers University of Technology (Drs. Hikansson and Carlsson), and the ENT Department, Sahlgren Hospital (Dr.Tjellstrom). Submitted for publication R b . 16, 1989; accepted Sept. 18, 1989. Reprint requests: Bo Hikansson, PhD, Department of Applied Electronics, Chalmers University of Technology, S-412 96 Goteborg, Sweden.
23/1/1m
known as the bone-anchored hearing aid)-which is coupled to an implanted part, the Brinemark System. The implanted part consists of a titanium fixture placed and integrated in the mastoid process, and a skinpenetrating titanium abutment.' Since 1977, more than 300 patients in Sweden and 200 patients in other countries have been provided with this type of hearing system, and numerous reports have been published presenting the surgical technique,* skin reaction^,^.^ and audiologic data.'-I0 Our experience with a percutaneous coupling to the bone is that in 92% of the cases," patients with a sensorineural hearing loss down to 45 dB PTA (pure tone average 0.5, 1, 2, 3 kHz) could benefit from the HC 200 sound processor when compared with their old hearing devices. Moreover, speech discrimination tests performed with background noise show that the HC 200 sound processor has a significantly better noise endurance for equal intelligibility than conventional bone conduction hearing aids. The credit for making hearing through bone conduction accepted in North America belongs to Dr. Jack Hough, MD, of Oklahoma City, and his co-workers. This group has developed a transcutaneous bone conductor, the Audiant device (Xomed Inc., Jacksonville, Ha.), which allows the skin and soft tissues to remain intact. The subcutaneous part of this device is a rareearth magnet, housed in a titanium disk, sealed and 339
340
HkANSSON et al.
OtolaryngologyHead and Neck Surgery
aim of this article is to theoretically-as well as experimentally-describe and discuss what effects these differences between the two systems can have on sound transmission quality and power requirements. TECHNICAL BACKGROUND Transducers for Hearing Aids
Fig. I.Schematic drawing illustrates components and parameters of fundamental importance in the performance of a variable reluctance transducer.
potted in medical grade silicone, and attached to the temporal bone by a Herbert orthopedic screw. Such a device was first implanted in the mastoid process above the sinodural angle in a patient in 1984. In 1986, the first paper was published describing the technique and the hearing results from the first 10 patients.I2 Among the audiologic criteria for implantation were bone conduction thresholds no greater than 30 dB PTA (0.5, 1, 2 kHz) and at no single frequency greater than 40 dB. In the discussion of that paper, the authors state that one of the major problems is associated with the power needed. At the American Academy meeting in Chicago in 1987, Dr. Houghi3stated (in his instructional session) that the bone conduction thresholds should not exceed 25 dB PTA and that the patients had difficulties excluding background sound from the subject of interest. Since 1984, approximately 500 patients in the United States and 100 patients in other countries have been provided with this type of bone conductor.i4 Although the basic principles of the transducers are the same (variable reluctance) in the two hearing devices, there are great differences in performance, as indicated earlier. The differences in performance between a percutaneous and a transcutaneous transducer are probably explained by the great difference in length of gap in the magnetic circuit, and the difference in suspension properties. The length of gap and the suspension properties are of fundamental importance for the performance of a variable reluctance transducer. The
Output transducers for hearing aids are named differently, depending on their applications. In airconduction hearing aids, receiver or earphone are the most appropriate names, whereas in bone-conduction hearing aids, names such as vibrator, bone conductor, or transducer are commonly used. For our purposes, the name transducer is used, and the discussion is restricted to transducers for bone conduction hearing aids. In the design of electromechanical transducers for bone conduction hearing aids, a solution that combines properties such as small size, wide frequency range, high output capability, low level of distortion, and low consumption of current is of utmost importance. Piezoelectric and magnetostrictive transducers are generally considered inconvenient, basically because of poor low-frequency response. Electrodynamic transducers of the moving coil type, which are most common in ordinary loudspeakers, are not considered becausefor physical reasons-they have too low an electrical input impedance, which means they require relatively high current under ordinary working conditions. Ordinary hearing aid batteries are not designed for such working conditions and, moreover, high consumption of current causes severe voltage fluctuations that, as a secondary effect, cause internal electrical stability problems. But an electrodynamic transducer of the variable reluctance type combines properties such as small size, wide frequency range, low levels of current consumption, and excellent electroacoustic performance, which are desirable in hearing aid applications. Prlncipal Design of a Variable Reluctance Transducer Principle of operation. In the following material, some equations are presented showing the influence of length of gap and of suspension properties on the performance of a variable reluctance transducer. These equations are described in greater detail by Carlsson and Hiikansson.l5 As indicated earlier, electromechanical transducers of the variable reluctance typeschematically illustrated in Fig. 1-are commonly used in conventional bone conductors. The mass M, including the coil, forms the counterweight. This mass should include a permanent magnet that produces the static magnetic flux $o. It should be noted that for a class A output amplifier, normally used in hearing aids
Volume 102 Number 4 April 1990
Percutaneousvs. transcutaneous transducers 341
with low-power capability, the static magnetic flux could be wholly or partially produced by the electrical current drain. The total magnetic flux is the sum of the static flux +o and the signal flux produced by the electrical current i. It can be shown that the total force F between the two sides of the gap is proportional to (denoted a ) the square of the total magnetic flux, yielding the relation
+ +-
F (Y
+ 2cu,+- + + -’= F, + F- + Fd (equation I )
under the assumption that the leakage of magnetic flux is negligible. From equation 1, it is apparent that the total force is composed of three forces: the static force Fo, the signal force F-, and the second harmonic distortion force Fd. Moreover, it is clear from equation 1 that it is important to have a static flux that is high in comparison with the signal flux, in order to achieve a low relative second harmonic distortion (Fd/F- a +-/+o).
Influence of length of gap on electromechanical conversion. The gap 6, as defined in Fig. 1, is the length of the magnetic path consisting of nonmagnetic material. Normally this gap is air, but it could also consist of some combination of biologic and inorganic material of low permeability, such as brass or various permanent materials. The magnetic flux +o is proportional to the strength of the permanent magnet (i.e., size and BH-product) and reciprocal to the gap 6 according to equation 2, if it is assumed that all magnetic reluctance is concentrated to the gap and that leakage is negligible. Equation 2 is stated as cb0 a (strength of magnet)/S (equation 2).
The electromechanical conversion from the electrical input current i or input voltage v to the mechanical signal output force F- is of the utmost importance and determines the maximum output capability, as well as the consumption of current for a given signal force level. The signal flux is proportional to the input current i, but is reciprocal to 6 as follows:
+-
+- a i / 6
(equation 3 ) .
Combining equations 1, 2, and 3, we find the following relation for the signal force as a function of input current and gap, assuming the strength of magnet to be unchanged: F_ a i1S2
(equation 4 ) .
From equation 4 it is clear that a large 6 can be devastating for the electromechanical conversion efficiency. As the average consumption of current determines the battery consumption, a large 6 results in high
TlQJ&@
Fig. 2. The percutaneous transducer system: Skull bone (I),soft tissue (2).titanium fixture (3), titanium abutment (4), bayonet coupling (5Jand percutaneous transducer (6).
battery costs. Moreover, it can be seen from equation 4 that an equal signal force for various gaps 6 is obtained if the current i is changed by the same amount squared (i a 6*). Combining equations 1, 2, and 3 yields that the second harmonic distortion is proportional to the gap squared: Fd a 6’ . F?
(equation 5).
Hence large gaps lead to high levels of second harmonic distortion that can be very annoying to the patients. In hearing aids, the maximum output capability is normally determined by the battery voltage. The input current i in equation 4 can be transformed to the input voltage v by the reciprocal of the electrical input impedance, according to Ohm’s law. Assuming that the electrical input impedance is determined by the inductance only (electrical resistance and hysteresis losses are disregarded), and recalling that the inductance is reciprocally proportional to the gap 6, then the signal force yields F- a v/6
(equation 6 ) .
For a given battery voltage, the maximum signal force can be considerably reduced because of a large gap 6. Influence of suspension properties on mechanical transmission. The signal force created in the gap is transmitted through the mechanical components of the transducer to the load. The suspension spring, characterized by its compliance C and inherent damping R,
OtolaryngologyHead and Neck Surgery
342 WkANSSON et al.
gap therefore requires a suspension spring with wellcontrolled properties.
Fig. 3. The transcutaneous transducer system: Skull bone (I), soft tissue of reduced thickness (21, internal part of transcutaneous transducer (57, and external part of transcutaneous transducer (4).
in combination with the transducer mass M, determines the basic resonance frequency and damping of the transducer. This particular frequency and this particular damping are of fundamental importance for the electromechanical performance of the transducer because they determine the frequency region, which is given an enhanced electromechanical sensitivity. In the following material, it is assumed that the mass M is the same in the two systems discussed. The simplified relation for the resonance frequency f is fa1 1 s
(equation 7 ) ,
if we assume that the load impedance is much greater than the output impedance of the transducer, which is the case in direct bone conduction, according to H%ansson et al.5 Furthermore, the mqimum output F, from the transducer is determined by the damping coefficient a according to the following equation: 1 F,a-a-
a
1 RvZ
(equation 8 ) .
Equations 7 and 8 show that the dynamic behavior of a variable reluctance transducer is strongly influenced by the suspension properties, compliance C and damping R. The suspension is also very important from the static point of view. The force created by static deflection of the suspension spring counteracts the attraction force from the permanent magnet and, at a certain length of gap, equilibrium is established. A narrow and stable
Percutaneous vs. Transcutaneous Transducers The percutaneous transducer. In Fig. 2, the principle of a percutaneous transducer is shown. A titanium fixture is permanently implanted into the mastoid bone of the skull and a permanent skin penetration is made with a titanium abutment. A bayonet coupling is used for attachment of the percutaneous transducer. The airfilled gap of the transducer is adjusted to an appropriate length, and the resonance frequency and damping (the region of enhanced electromechanical sensitivity) are chosen to fit the amplification need for the typical hearing loss among the patients this hearing device is aimed for by selection of appropriate mass M, compliance C , and damper R. The transcutaneous transducer. In Fig. 3, the principle of a transcutaneous transducer is shown. The implanted part of the transcutaneous transducer is housed in biocompatible material and secured to the mastoid bone of the skull. The permanent magnet (or magnets), preferably housed in both the external part and the internal part of the transcutaneous transducer, not only have the purpose of reducing second harmonic distortion, but should also keep the device safely in place over the intact skin and soft tissue. For this transducer principle, the gap is filled with the skin and soft tissues of reduced thickness, the properties of which then determine the length of the gap, as well as the resonance frequency and the damping. DISCUSSION
The purpose of this article has been to discuss some of the fundamental properties of a variable reluctance transducer with reference to percutaneous vs. transcutaneous bone conduction. Transcutaneous transmission means that there is no metallic contact between the internal and external parts of the transducer-instead, there is intact soft tissue in between. The thickness of the soft tissue determines the gap length, which-in this situation-is relatively large. With a percutaneous transmission system, on the other hand, the gap between the two parts of the transducer is small and kept constant by a metallic spring suspension. It is apparent that, with certain assumptions, large gaps can be devastating for power consumption and maximum output capability, as well as for second harmonic distortion. One aspect that can influence these results is that the strength of the magnet is assumed to be the same for the two systems discussed. It appears from equations 1 and 2 that a strong magnet can solve most of the problems associated with large gaps. This is misleading, however, because an increased amount of permanent magnet material will significantly reduce
Volume 102 Number 4 April 1990
Percutaneous vs. transcutaneous transducers 343
100
200
500
lk Frequency Hz
2k
5k
1Ok
Fig. 4. Output force levels at 50 d B SPL input (lower broken and solid lines) and at saturation (upper broken and solid lines), produced by t h e Audiant A.T.E. sound processor (solid lines) and the HC 200
sound processor (broken lines). the dynamic transmission efficiency, since high quality permanent magnet materials (such as rare earth cobalt) have poor dynamic properties. More details of these aspects were given by Carlsson and Hakanss~n.'~ Also, the strength of the permanent magnets is closely related to the size (or weight) of the magnet. A large size is not desirable for aesthetic reasons, and a heavy weight leads to problems keeping the transcutaneous transducer in place. One assumption in the analysis was that the magnetic leakage flux was negligible. This is true for narrow gaps, but for wide gaps a portion of the signal flux passes outside the area of the gap; this implies a reduced dynamic transmission efficiency. The length of the gap, as well as the properties of the soft tissue maintaining the gap in a transcutaneous transducer, will vary individually. The tight scar tissue layer of an adult who has been exposed to previous surgery might have mechanical characteristics different from the tissue of a child who has not been operated on before. There could also be changes over time in the individual patient. From the use of conventional bone conduction hearing aids, it is well-known that the patients often will have impressions of the soft tissues because of the pressure needed. Many patients using such conventional bone conduction aids cannot have the transducer in the same position all day and will move it around and even take it off during certain periods. The force needed for a transcutaneous device will probably be of the same magnitude, and some of the
same problems could be anticipated. In conclusion, one might expect that the transcutaneous device will give relatively high intersubject as well as intrasubject variation in electromechanical transmission in comparison with the percutaneous device. The percutaneous transmission system has, of course, the potential disadvantage of inflammation and infection at the site of the skin penetration. The experience with percutaneous implants in the mastoid, however, goes back to 1977, and the frequency of skin problems has been low.' More than 360 patients have received a total of about 580 implants in the mastoid process for a direct bone conduction instrument or for retention of auricular prostheses. These patients have been followed carefully and the frequency of adverse skin reactions has been registered. More than 3000 observations have been made, and in 90% of those no adverse reactions of the skin around the skin penetration were noted. About 4% of the observations were of a type that called for extra visits and treatment. Only one of 250 implants for direct bone conduction had to be removed as a result of soft tissue problems. Preliminary Experimental Comparison
There are differences between the percutaneous and the transcutaneous hearing systems in terms of how to perform objective (patient-independent)quality control of the complete hearing device. For the percutaneous device, a separate unit, the skull simulator TU-1000,
344
&ANSON
OtolaryngologyHead and Neck Surgery
et al.
described by Hakansson and Carlsson,I6 has been developed. The skull simulator is capable of measuring the output force level from the HC 200 sound processor, with a load impedance relevant for the impedance of the normal patient. This measurement procedure cannot be carried out with the transcutaneous device in a similar way because a vital part of that transducer is totally implanted. In order to perform comparative experiments between the Audiant A.T.E. sound processor and the HC 200 sound processor, however, an adaptor was developed so that an implantable orthopedic screw of the Audiant device could be rigidly attached to the skull simulator TU 1000. The internal and the external units were separated by application of a 1.5 mm-thickness skin transplant in between. Results from measurements of the output force level produced by the two devices are presented in Fig. 4. The measurements were performed with all external adjustable controls set to produce maximum output force level. From the saturation force levels (SFL) shown in Fig. 4 by the upper curves, it was found that the HC 200 sound processor could produce approximately 20 dB higher maximum force levels than the Audiant device. This result is in agreement with the great difference in maximum allowable bone thresholds (45 vs. 25 dB PTA between the two hearing systems). The sound pressure level for the measurement of the linear frequency response had to be lowered from 60 dB, which is normally used for hearing aids, down to 50 dB to achieve sufficient linearity of the Audiant device. From the output force levels recorded at 50 dB (OFL,,) and the SFL it seems that the maximum dynamic range is considerably smaller for the Audiant device than for the HC 200 sound processor. The difference in consumption of current was measured with pure tones at l .O, l .6, and 2.5 kHz, respectively, and at an equal output level for the two devices when working in their linear range. It was found that the Audiant device (as an average over the three frequencies) required 26 times higher current than the HC 200 sound processor for an equal output force level. CONCLUSION
There is basically one reason to choose the transcutaneous device in preference to the percutaneous one-namely, to avoid permanent skin penetration. On the other hand, the percutaneous device offers the opportunity to optimize the transducer for a stable and well-controlled performance, and, above all, to reach high efficiency in electromechanical transmission. In view of the low complication rates for the permanent skin penetration and fixture implant, and the superior electromechanical performance, a percutaneous trans-
ducer system is chosen in preference to a transcutaneous one for the bone-anchored hearing aid, also known as the Nobelpharma Auditory System HC 200. We wish to acknowledge the skillful technical assistance of research engineer Berndt Anderson in the accomplishment of the experimental session. REFERENCES
1. Tjellstrom A, Lindstrom J, Halltn 0, Albrektsson T, Brinemark P-I. Osseointegrated titanium implants in the temporal bone. A clinical study on bone-anchored hearing aids. Am J Otol 1981;2:304-10. 2. Tjellstrom A. Osseointegrated systems and their application in the head and neck. Adv. Otolaryngol Head Neck Surg 1989; 3:39-70. 3. Holgers K-M, Tjellstrom A, Erlandson B-E, Bjursten LM. Soft tissue reactions around percutaneous implants: a clinical study on soft tissue conditions around skin penetrating titanium implants for bone-anchored hearing aids. Am J Otol 1988;9:56-9. 4. Jacobson M, Tjellstrom A, Tomsen P, Albrektson T. Soft tissue infection around a skin penetrating osseointegrated implant: a case report. Scand J Plast Reconstr Surg 1987;21:225-8. 5. Hikansson B, Carlsson P, Tjellstrom A. The mechanical point impedance of the human head, with and without skin penetration. J Acoust Soc Am 1986;801065-75. 6. Hikansson B, Tjellstrom A, Rosenhall U. Acceleration levels and thresholds with direct bone conduction versus conventional bone conduction. Acta Otolaryngologica (Stockh) 1985;100:24052. 7. H&ansson B, Tjellstrom A, Rosenhall U. Hearing thresholds with direct bone conduction versus conventional bone conduction. Scand Audiol 1984;13:3-13. 8. Carlsson P, Hikansson B, Rosenhall U, Tjellstrom A. A speech reception threshold test in noise with the bone-anchored hearing aid a comparative study. OTOLAFXNGOL HEAD NECK SURC 1986;94:421-6. 9. Tjellstrom A, Jacobson M. A 10-year experience with maxillofacial bone-anchored implants. Presented at the Annual Meeting of the American Academy of Otolaryngology-Head and Neck Surgery, Washington, D.C., Sept. 25-29, 1988. 10. Tjellstriim A, Jacobson M, Norvell B, Albrektsson T. Patient attitudes to the bone-anchored hearing aid: Results of a questionnaire study. Scand Audiol 1988;18:119-23. 11. Hikansson B, Lidtn G, Tjellstrom A, Ringdahl A, Carlsson P, Erlandsson B-E. Ten years of experience of the Swedish boneanchored hearing system. Ann Otol Rhino1 Laryngol: submitted 1989. 12. Hough J, Vernon J, Dormer K, Johnson B, Himelick T. Experiences with implantable hearing devices and a presentation of a new device. Ann Otol Rhino1 Laryngol 1986;95:60-5. 13. Hough JVD, Dormer KJ. Implantable hearing devices: XOMED bone conductor. American Academy of Otolaryngology-Head and Neck Surgery, Course 3337, 1987. 14. Campos CT. A chronology of an implantable bone conductor hearing device. Hear Instrum 1988;39:36, 38, 78. 15. Carlsson P,Hikansson B. A new transducer for hearing by direct bone conduction. Technical Report No. 2:88. Research Laboratory of Medical Electronics, Chalmers University of Technology, Goteborg, Sweden, 1988. ISBN 91-7546-050-5. 16. Hikansson B, Carlsson P. Skull simulator for direct bone conduction hearing devices. Scand Audiol 1988;18:91-8.
