With ia the auditory assistants improve to amplify the voice and block noise

With the purpose of making listening more clear and pleasant for people who suffer from ear loss and use auditory assistants, artificial intelligence (AI) has begun to use investigations in order to improve those devices. In fact, there are already intelligent hearing aids in the market and in the case of cochlear implants, studies are advancing a lot, said Gonzalo Corvera Behar, founder and director of the Mexican Institute of Otology and Neurotology (IMON).

These auxiliaries amplify the sound, although that function does not solve the problem of the person who uses them; They help, make sounds perceive, but these are mixed with the noise and the voice is not clear. You have to teach the device to highlight it. The AI ​​is used to generate the algorithms that will allow the device to know what is noise and what is voiceaccurate.

The specialist added: It is not that the device uses artificial intelligence continuously, because that would take too much time and should work instantaneously. If it takes more than three milliseconds to process the sound, the patient will not have the feeling of hearing naturally, he will hear with echo. Hearing loss is never a couple, that is, more frequencies are ending more than others and if those that are well listening to before those that are wrong, they sound like an echo and it is unpleasant.

Machine learning (machine learning) It is the technology that deals with these research. Allows equipment to learn data. There are several aspects of her, but one is that instead of wanting to program the computer, different ways of reacting are given and we look at the one that produces the result we are looking for. Basically, it is a way of gathering user informationhe detailed.

Digital processing

In the case of cochlear implants, it goes advancing, although we are stuck in knowing what is the optimal way to create the interface between them and the auditory nerve, that is, what is the digital processing that the artificial ear must do to pass the signal to the nerve. The study consists of connecting patients implanted to a computer by which a lot of parameters is obtained that is possible to change and see the gain. We inquire about which it is ideal to modify in each personheld.

“Likewise, we send to the patient exercises in a kind of test, which will tell us how well listening; The result is captured by the computer and we compare it with the way in which the implant is programmed, and together the information of thousands and thousands of cases we can know that it is working better.

I think that around cochlear implants AI has a lot of action field. From very practical things, such as reducing the time that a patient’s intervention takes on us, but above all improve the result. He considered that in the long run there may be devices that are continuously connected to the cloud and learn how to adjust.

What is sought is that the device can know what we want to pay attention, to amplify the voice of the person with whom it is said, delete those around those around; For example, in a restaurant, who knows what we pay attention to and just listen to that. It seems of science fiction, but there are already some experimental systems that are managing to do so. It is about reading the mind in order to adjust the device. However, this will not come the year that entershe said.

When referring to Tinnitus, he explained that there are several strategies to reduce that condition. One is masking, in which we can use AI to determine which are the most useful parameters. In this problem we rather focus on measures that may seem psychological, since it is a brain response: in the absence of sound that comes from around, that organ applies its own noises. We seek to train it and do not give them importance.

For people who begin to lose their hearing, he recommended making their annual review, Because the only way we can give them the best technology is to revalue their case. It is worth maintaining our ability to communicate with other people

Deafness and dementia

He assured that the relationship between deafness and dementia is very clear. If someone suffers superficial loss at 50 or 60 years, 15 or 20 years later they have twice the chances of suffering from it; If it is medium grade, the risk is triple, and if it is severe it is five times more likely to suffer that condition.

Corvera concluded: One of the factors that cause that damage is that the people who do not hear are isolation, do not participate in the conversation, so their mind is less active. Another is that the person is making a great effort to listen to what others say, so it is under a state of stress and it causes the slightest regeneration of nerve cells.