Improving Speech Intelligibility

Professor Miriam Furst-Yust from Tel Aviv University in Israel, has developed a new type of ear implant that can improve hearing for the hearing impaired. The device makes it easier to single out and understand speech in a noisy environment.

Hearing impaired individuals depend on technology to function normally in everyday life. For healthy people, the hearing organ contains ten thousand tiny outer hair cells (OHC). Damage to the OHC is one of the leading reasons for a decline in hearing. The hair cells move in correspondence to the mechanical wave a sound creates in the air and transfer electric impulses to the vestibulocochlear nerve. This nerve transmits the auditory information to the brain where it’s processed further and is translated into meaningful signals.

Existing hearing aids use the basic biology of the ear to reproduce the sound. A common solution to a hearing problem that results from hair cell damage is placing an implant into the cochlea. Cochlear implants covert the mechanical wave that reaches them to digital information which is then translated to electrical pulses which are passed on to the vestibulocochlear nerve. Most of the implants’ technology is based on wave frequency conversion. They predict the reaction of the cochlea, but do not always simulate correctly the properties of cochlear signal processing, especially when complex signals such as fluent speech are involved. While existing hearing aids have proven themselves efficient in a relatively noiseless environment, they fail to provide a good hearing experience in a noisy one.

Professor Miriam Furst-Yust has developed a new algorithm for sound conversion based on a mathematical model that better simulates the process that occurs in the ear once a sound wave reaches it. Her team incorporated an outer hair cell model which acts as an amplifier and increases the cochlear response. This affects mostly the middle part of the cochlear partition, which corresponds to the characteristic frequencies of 2 to 6 kHz. The team effectively replaced the organ which normally works as a sound amplifier with a mathematical model that analyses the mechanical wave and has the same functionality as the OHC. On the basis of the cochlear response, the researchers have developed a heuristic algorithm for reconstruction of noisy speech signals.

Implementing this new algorithm should not be a difficult task for hearing aid manufacturers, as the alternation requires only minor re-engineering. Once the alteration is done, the invention can improve the quality of life for many hearing impaired individuals worldwide.

TFOT has recently covered a computerized lip reading system developed at the University of East Anglia, which was designed to help hearing disabled individuals. We have also covered the discovery of hair cells that have an important role in transmitting molecular signals, made by researchers from Yale University.

For more information on the novel technology, please visit the Ramot site for Tel Aviv University technology patents.