Diagnosis of Covid19 by AI analyzing the voice

in Project HOPE4 years ago


Source

We have all heard of the use of artificial intelligence to combat the virus in question, analyzing data, comparing radiographs of the lungs or PCR tests.

But, if this does not surprise us anymore, the latest application developed on artificial intelligence is capable of detecting if we are infected, only by analyzing our voice, even if we still do not have any symptoms of the disease.

This technology is not totally innovative, since it had previously been used to detect other pathologies such as heart failure or heart attacks, which, it seems, also affect the tone of voice.


Source

According to its developers, if, due to the disease, it is difficult for you to breathe, it is logical to think that it also affects you when speaking, this produces a change in the wavelength of the voice and can be detected by artificial intelligence.

In this situation, we only have to prepare the appropriate neural network and supply thousands and thousands (the more the better) of voice recordings with the label of whether or not the speaker is infected and give the neuronal network the appropriate training.

After training the neural network, by analyzing a new recording, it will be able to tell us whether or not the owner is infected with coronavirus.


Source

The truth is that, if we leave confinement, deaths and economic ruin aside, we have to recognize that, thanks to or because of this damn bug, some technologies will advance in a year, which would normally take us 5 years or more.

And I am not only referring to the health field but also in regards to the economic, labor and social issue, such as the issue of teleworking that is changing the lives of many people.

I know that this is not a consolation for most people but I try, at least, to find the positive side of all the rubish that we are suffering.


gb_peq.png spanish-flag.png
Versión en español

Coin Marketplace

STEEM 0.19
TRX 0.15
JST 0.029
BTC 63239.86
ETH 2621.03
USDT 1.00
SBD 2.77