Cochlear Implants are one of the oldest types of sensory neural prosthetics in existence. They are designed to replicate, as best as is possible, the sense of sound. They transmit to the brain, a series of patterned electrical impulses which the brain is able to learn to recognise and process as sounds, making out words and syllables from a silent world.

However, there are problems with it.

Cochlear Implant

The Cochlear is a sealed, spiral container of fluid. It is almost 3cm in length, but half that length is the semicircular canals, also sealed, that are part of the balance mechanism, and can be discounted. In the remaining length of the Cochlear, including the entire spiral, is the key to the sound detection mechanism - anywhere between 16,000 and 20,000 tiny white hairs. Everything up to this point has just focused and amplified incoming vibrations - the Cochlear is where the sound is actually detected.

Each hair follicle is microscopically different in length to those around it. This means that each is attuned to a slightly different frequency of vibration, and can pick up and send a signal for that frequency only. There is more than one hair at any given length, but only a very small number. Those that do share the same length, for any given length, have differing degrees of resiliency to the fluid which passes over them. A hair with a low resistance will pick up even a very faint sound at that frequency, whilst a hair with a high resilience, will only pick up a very loud sound. This feature is highly useful, not just at detecting loudness, but because repeated exposure to over-loud noise damages, then destroys the sensitive hairs, and they do not grow back once destroyed.

The original Cochlear implants had just one electrode. This meant they triggered an almost infinitesimal share of the user's auditory nerves. This meant the very luckiest users managed to pick out about 5 percent of words spoken, without using lip reading. Common words like 'lot' and 'home', 'yes' and 'no' were hit and miss at telling apart.

Initial implants were placed against the outer wall of the inner ear, reverberating through the Cochlear's outer walls. The problem this caused was the electrodes could easily slip, resulting in a complete loss of function. An improvement in the hardware was sought, and it was finally decided to try to drill into the Cochlear, and implant electrodes within the fluid chamber itself. This proved successful, and had the added benefit of preventing the electrodes from slipping. Unfortunately, a side effect of perforating the Cochlear, is all residual natural hearing ability in that ear is usually lost, leaving the electrode array as the only means of viable hearing.

Worse, these implants have a limited range of tones because it is difficult to insert the electrode array beyond the outer turns of the cochlea. The outer turns pick up the high frequencies, so people using today's implants are sensitive only to the highest tones.

The implant in the Cochlea

We finally understand enough about the way sound signals are processed into electrical signals, to go one better than the cochlea implant. We can tap directly into the auditory nerve itself.

All the hairs in the cochlea have their own nerve channels, which all connect together as one large nerve leaving the ear, called the auditory nerve. It is to this nerve directly that a Cochlear implant connects. The signals from sixteen thousand or more natural hair cells travel up this nerve at a time. There is no way a 24 electrode implant could compete.

Now, John Middlebrooks of the University of Michigan, US with Russell Snyder of the University of California, San Francisco also US, has tested on cats, a prototype direct aural nerve connection.

Experiments in cats that compared the brain's response to the established designs created by the tone signals from a natural cat ear with its response to the prototype show the new device dramatically improves the range of tones that can be heard.

"Current implant users do very well with speech in a quiet environment but struggle with background noise," Middlebrooks explains, "they also have very poor pitch perception and cannot appreciate music - in an environment like a crowded room, you use pitch to tune into a person's voice."

Experiments on 10 cats involved first recording the response of the brain to a range of tones. The cats were then deafened and the same tones were played with a conventional, and then experimental, implant installed.

Conventional implants only allow detection of tones as low as around 7 kHz, but the new implant allowed frequencies as low as 0.6kHz to be detected.

The new design also produced a cleaner/more specific response in the brain to a particular frequency, and used much less power to get results. "If we develop this for humans it would have much lower power demands, so it could be much smaller and perhaps fully implanted," says Middlebrooks.

Tests in which animals are fitted with the implants for periods of months are now planned.

"Cochlear implants themselves have provided a revolution," says Brian Lamb, director of the Royal National Institute for the Deaf. "These implants - if successfully transferred to people - could offer further, major benefits."