Background noise eliminated with hearing aid breakthrough

Background noise eliminated with hearing aid breakthrough
A new type of hearing aid is being developed by researchers, designed to filter out background noise and render speech easier to understand to allow hard of hearing users to follow a single speaker.

People with hearing difficulties can be challenged to follow conversations in loud environments with multiple participants, being unable to use the ‘cocktail party effect’, the ability to focus on individual speakers and filter out background noises. 

Existing hearing aids struggle to establish a link between the ear and the brain that accounts for selective hearing as Dr. Axel Winneke, research fellow, Fraunhofer Institute for Digital Media Technology (IDMT), said: “For people with normal hearing, the connection between the ear and the brain is unimpaired.

"This is how they know which direction to face whenever they hear a signal. For hearing-aid wearers, this ability is severely restricted. Even high-end devices are unable to pinpoint the source of an acoustic signal of interest. We therefore need to retrieve this information from the brain. Using an electroencephalograph (EEG), we can analyse the brain activity of a hearing-impaired person and determine who it is they are listening to.”

The EEG analysis was conducted by Fraunhofer IDMT-HAS and the University of Oldenburg, with Winneke and other researchers currently working on a system which is designed to improve speech intelligibility in busy environments. 

The technology is based on a combination of EEG, audio-signal processing and electrostimulation of the auditory cortex, with the EEG measuring brain activity to determine the direction in which the user is trying to hear, creating an interface between the brain and a microprocessor. 

Information is then transmitted to the hearing aid, focusing a beam-forming microphone in the intended direction of hearing, amplifying the specific audio signal that the hearer is attempting to listen to.

The audio signal is focused, filtering out background noise sources such as other voices, with transcranial electrostimulation employing a low electrical current to stimulate the auditory cortex on the basis of a voice signal, improving the intelligibility of talking people. 

The stimulation process and requisite hardware are being developed by neurConn in partnership with the University of Oldenburg.