Hearing impairment affects an estimated 10% of the population.  About 1 in 1000 babies are born profoundly deaf, another 1 in 1000 becomes deaf before adulthood, and half of people above 60 years old suffer from hearing loss. (1994)

The anatomy of the ear is shown in the figure on the left.  The ear consists of the outer, middle and inner ear, and hearing loss is classified based on the location of damage.  Damage to the middle ear is called conductive loss, and damage to the inner ear is called sensorineural loss.  When conductive hearing loss occurs in combination with sensorineural loss, it is referred to mixed hearing loss.

Anatomy of the ear

The degree of hearing loss is described as mild, moderate, severe or profound.  For over a century, hearing aids have been the primary treatment option for sensorineural hearing loss.  Conductive hearing loss can often be medically or surgically corrected, but some opt for hearing aids instead.

During the past few decades, implantable auditory prostheses have been developed, including: 1) cochlear implants for the profoundly deaf with sensorineural impairment which stimulate the eighth nerve directly through electrodes positioned within the cochlea; 2) direct bone-conduction devices which transfer sounds via bone conduction to the cochlea, bypassing the damaged middle ear; and 3) devices with a piezoelectric or electromagnetic transducer that directly drives one of the middle ear ossicles.  This report will focus on the third class of implantable auditory prosthesis, known as “middle ear implants” (MEIs) or “soundbridges”, for sensorineural hearing loss.(Snik, Mylanus et al. 1998)

There are several problems with conventional hearing aids:

First, the feedback (annoying whistling or squeal) produced by hearing aids can be a problem. This is due to amplified sound traveling back to the input microphone, resulting in feedback. As the degree of amplification required increases and as the size of hearing aid decreases, the problem of feedback increases. People with severe hearing loss are precluded from wearing tiny hearing aids due to the possibility of feedback.
	Second, ear canal discomfort is a problem for some people, due to tight fitting earmolds. These earmolds can cause irritation, allergies and infection.
Third, the occlusion effect, in which one's own voice sounds different due to occlusion in the ear, is a serious and continuing problem for some hearing aid users. Because of the occlusion, one's own voice is perceived primarily through bone conduction, in contrast with the unoccluded ear wherein perception is through both air and bone conduction. The effect is that the user's own voice sounds different, which some people describe as "funny", "hollow", "booming", like talking in a barrel, like hearing echoes of their own voice, etc.
	Fourth, distortion is caused by anatomical limitations and placement of the hearing aid. At low amplification levels, the distortion is minimal, but at high amplification levels, the distortion can be significant due to the cavity between the hearing aid and eardrum in which sound resonates.
Finally, there is often a perceived stigma associated with wearing hearing aids. The less visible MEI may be more readily accepted by people who feel that wearing hearing aids is associated with a handicap and getting older.
Hearing aids	Middle ear implant

Advances in technology, such as the reduction in size and improved reliability of electronic components and our increased knowledge on implantable materials and implantation techniques, have also helped propel the development of MEIs forward.(Snik, Mylanus et al. 1998)

MEIs have been around since 1935 when Dr. Wilska sprinkled iron fillings into a person’s eardrum, creating a magnetic field when coupled to an external coil of wire.  Subjects reported hearing from an earphone that sends its output through a coil of wire instead of transducing electric energy into sound.  Iron fillings, acting as an internal receiver, vibrated in synchrony with the magnetic field, causing the eardrum to vibrate.  Sound is then transduced normally in the subsequent pathway to the inner ear.  The first experimental device was bulky in size and power-consuming, requiring 28000 mA to produce 85dB SPL and iron fillings were not fixed on the tympanic membrane and proper operation depended on the person’s anatomical position.(Chasin 2000)

MEIs have evolved over the years, and today MEIs are wearable and power-efficient, requiring only 3mA to produce 85dB SPL.  Historically, MEIs were first clinically available to those with unresolvable middle ear conductive/mixed losses (e.g. broken ossicular chains).  However, modern MEIs are designed for those with sensorineural hearing loss and require a well-functioning ossicular chain.(Chasin 2000)

There are two types of MEIs – piezoelectric and electromagnetic.  The piezoelectric approach， pioneered by Drs. Suzuki and Yanagihara, uses a piezoelectric crystal, which has an interesting property– applied electric charge can cause the crystal to bend and a bent crystal generates electric charge.  The crystal can function as a microphone, generating electric charge in response to incoming sound waves which bend the crystal, and as a driver (when attached to the middle ear bones), moving in response to electric charge from the microphone.  This causes the middle ear bones to vibrate and transduce sound to the inner ear.(Snik, Mylanus et al. 1998; Chasin 2000)

The electromagnetic approach uses an external microphone and sends the signal through an inductive coil that creates a magnetic field.  The implanted receiving coil picks up this signal and connects to a transducer attached to one of the three bones and vibrates in synchrony with the magnetic field.  Sound is then transduced to the inner ear.(Snik, Mylanus et al. 1998; Chasin 2000)

One advantage of the piezoelectric approach is that the components of the implant are physically small, allowing a fully-implantable system.  Its design is also simple.  In contrast, the electromagnetic approach uses bulkier components and are only partially implantable.  Its design can also be complex.  However, compared to the piezoelectric approach, the electromagnetic approach can provide significantly more gain and output and can be used for patients with severe hearing loss.  The piezoelectric approach is limited to patients with mild to moderately severe hearing loss due to its limited output gain.(Snik, Mylanus et al. 1998; Chasin 2000)

FDA-Approved Devices

At present, there are two FDA-approved middle ear implants on the market: (Clabo 2002)

Vibrant Soundbridge (MED-EL) http://www.vibrant-medel.com/ http://www.fda.gov/cdrh/mda/docs/p990052.html Type: Electromagnetic partially implantable Cost: $15,000 (including surgery)
Direct Drive Hearing System (Soundtec, Inc) http://www.fda.gov/cdrh/mda/docs/p010023.html Type: Electromagnetic partially implantable Cost: $5,000

Devices Under Clinical Trial

Still in clinical trials: (Clabo 2002)

• Envoy Middle Ear Implantable System (St. Croix Medical, Inc.)
  http://www.envoymedical.com/index.htm
  Type: Piezoelectric fully-implantable
• Middle Ear Transudcer (Otologics LLC)
  http://www.otologics.com/us/index_us.cfm
  Type: Electromagnetic partially implantable

This section discusses the design of one electromagnetic partially-implantable middle ear implant currently available on the market -- MED-EL's Vibrant Soundbridge.

Symphonix’s Vibrant Soundbridge, shown above, consists of an external and an internal component.  The external part, called the Audio Processor (AP), is worn on the head underneath the hair and held with a magnet and the internal part, called the Vibrating Ossicular Prosthesis (VORP), is implanted under the skin.

The AP (shown above), available in different colors to match hair color, consists of a microphone, digital signal processor (DSP), external coil, programming socket and battery compartment.  The microphone picks up environmental sound, and the DSP processes the signal, by first dividing the input into several frequency bands and applying amplification algorithms such as linear amplification or wide dynamic range compression on individual bands to suit the individual's audiological profile. The output of all bands are summed and translated for mechanical vibration. The signal is then converted into RF energy for transcutaneous transmission through the external coil.  The battery compartment holds a conventional type-675 hearing aid battery, which lasts approximately one week with normal use.  The programming socket allows the audiologist to program the device according the patient’s hearing loss.

The VORP (shown above) consists of a magnet, receiving coil, demodulator, conductor link, and floating mass transducer (FMT).  The magnet allows the AP to be held in position over the implanted part.  The receiving coil is coupled with the external coil and the electromagnetic signal is demodulated and sent to the FMT via the conductor link.  The FMT, attached to the incus, converts electric signal into vibrational energy, which are then transmitted to the inner ear via the ossicular chain.

The components of a partially-implantable middle ear implant, such as the external microphone, external audio processor, transmitting and receiving coils, modulation and demodulation circuitry, and data packet design, can be adopted from cochlear implant technology. Reviews on these topics can be found in literature and the design of such components can be found in the patent database. The only difference is in the actuator -- the cochlear implant uses an electrode array to stimulate different places along the cochlea whereas the middle ear implant uses an actuator (the FMT) that directly drives the middle ear ossicles.

The FMT has a "floating mass", which vibrates in response to a signal corresponding to sound waves. Electrical signals are sent to two electromagnetic coils wound around a hermetically sealed titanium housing. A magnet inside the housing supported by a pair of springs vibrates in response to the magnetic field generated by the electromagnetic coils, causing the entire structure to vibrate.

The floating mass is mechanically coupled to a housing, which is directly mounted on one of the middle ear bones, which is the incus in this case. The device's operation is based on the principle of conservation of energy, which states that energy can neither be created nor destroyed, but can only be changed from one form to another. Specifically, in a connected system of bodies, mechanaical energy is conserved (excludidng friction), and if one body loses energy, a connected body gains energy. (Ball et al, 1999)

The figure below illustrates the basic theory of operation. Figure a below shows a floating block (FMT) connected to a counter block (incus) by a flexible connection. As the floating block vibrates in response to sound waves, mechanical coupling provided by the flexible connection allows vibrations to be transmitted to the counter block, which generally vibrates "counter" to the floating block. Figure b illustrates the "counter" vibration with arrows showing relative vibration of each block. The relative vibration of each block is inversely proportional to its inertia, which depends on mass and other factors (e.g. whether the block is connected to anoother structure). Figure b illustrates the case in which the relative inertia of two blocks are equal; in figure c, the mass of the floating block is greater than the counter block; and in figure d, vice versa. Case c is used when the FMT is designed to transduce mechanical vibration of the eardrum into electrical signals (e.g. as a sensor in a fully-implantable system (not yet available on the market) where the eardrum acts as a microphone); greater vibration of the counter block is desired for greater sensitivity. Case d is used in the Vibrant Soundbridge, where the FMT as an actuator transforms vibration into vibration of the incus. Having a floating mass with greater inertia also provides greater sensitivity.

The FMT can be mounted on any vibrational structure in the ear with a mounting mechanism such as glue, adhesive, screws, clip, etc. The figure a below shows the FMT attached to the incus by a clip, and figure b shows a perspective view of the FMT.

The FMT is designed to mimic the vibratory response of the middle ear, and is capable of delivering mechanical stimulation with a frequency range corresponding to the speech frequency range. Laser doppler anemometry has been used to verify the device's operation. Clinical studies show that the device is safe and effective, and patients could hear as well with the device as conventional hearing aids. Changes in residual hearing are clinically insignificant. There is also significant improvements across various listening modes for the majority of patients, and qualitative reports from patients demonstrate a positive reaction.

Today there are 28 million hearing-impaired Americans and only 6.6 million wear hearing aids.  The market for middle ear implants, consisting of patients who are dissatisfied with hearing aids or unable to benefit from hearing aids, can easily reach several million. Clinical studies have confirmed the benefits of middle ear implants over conventional hearing aids, but risks of surgery apply.

Having a partially implantable MEI has its advantages.  With the AP worn externally, patients can avail of improvements to signal processing technology without further surgery.  Symphonix’s AP, for example, has changed from a three-channel to an eight-channel device, allowing more flexibility in tailoring the device to a patient’s audiogram.

Future MEIs could be fully implantable, such as illustrated in the figure above, in which two FMTs are used -- one as a sensor to transduce eardrum vibrations and one as an actuator to make one of the middle ear bones vibrate. The electronics, sealed hermetically, are implanted subcutaneously. Improved fabrication / manufacturing technology could lead to smaller devices with even more features such as connectivity to consumer electronic products such as TVs, cellphones and MP3 players.

(1994). Deaf or Hard of Hearing Population Statistics, League for the Hard of Hearing. 2006.
           
Chasin, M. (2000). Middle Ear Implants. Audiology Online.
           
Clabo, K. (2002). Middle ear implants not your grandpa's hearing aid. CNN.com.
           
Snik, A., E. Mylanus, et al. (1998). "Implantable hearing devices for sensorineural hearing loss: a review of the audiometric data." Clin Otolaryngol 23: 414-419.

Otologics, LLC. http://www.otologics.com.

Vibrant Soundbridge. http://www.vibrant-medel.com/.

Ball, et al. (1999). Bone conducting floating mass transducers. US Patent # 5,913,815.