If there are algorithms artificial intelligence able to detect respiratory diseases from coughs, why not do the same with intestinal infections? Logically, it would not be the cough that would have to be recorded, but the bowel movements. Come on, we would have to install a recorder in the bathroom to distinguish between diarrhea and a normal consistency. All this may seem like a joke, but it is actually the purpose of an investigation of the Georgia Institute of Technologywhich has just been presented at the 183rd Meeting of the Acoustic Society of America.
Specifically, the objective of these scientists is detect diseases such as cholera. It is a common disease in places where the conditions of water sanitation they are deficient. It can occur because they are areas with few economic resources or in which some kind of natural disaster.
Thanks to this artificial intelligence algorithm, the study authors believe these areas could be tracked to detect outbreaks before they become a serious public health problem. And it is not a simple diarrhea. It is a disease that can be very serious, even fatal, especially because it occurs precisely in places where access to suitable drugs it’s more complicated. In the absence of medicines, cutting the shoots at the root is vital.
What is cholera?
Cholera is an intestinal disease transmitted by bacteria Vibrio cholerae. It is mainly transmitted by the consumption of undrinkable water or food contaminated with fecal matter.
Its main symptom is a very watery diarrhea, which quickly causes dehydration in affected people. Therefore, in places with difficult access to antibiotics, which are precisely the most affected, mortality can be very high.
Detect diarrhea using artificial intelligence
The authors of this new research have proposed the use of an artificial intelligence algorithm capable of detecting a possible cholera outbreak by recognizing the different types of depositions.
To do this, the first step is to place a recorder that captures sound in a way non-invasive for users. These sounds are then transformed into images through a technique known as spectrogram. This makes it possible to capture patterns associated with the different types of depositions. For example, the sound of urine is more continuous, while that of stool is usually much more concise. However, when there is diarrhea, it is more random and gives rise to some heterogeneous but recognizable patterns.
These spectrograms are passed through a artificial intelligence algorithm, which was previously trained by the researchers with recordings of known situations. Recordings with or without background noise were used, to ensure that the algorithm could precisely capture the desired information.
What is all this for?
The goal of these researchers is to place their device in bathrooms in places where sanitation problems cause habitual outbreaks of cholera. Also in areas where there have been recent natural disasters or which, for whatever reason, have a greater punctual propensity. Thus, they could detect the places where an outbreak begins to appear and cut it off before patients even begin to seek help.
That would be the main objective. However, these scientists do not rule out that in the future it could even be added to the functions of any smartphone, basically to control intestinal transit in a most curious way. That would be a little more unnecessary, because as users we can know if we have diarrhea or not, but we would save having to enter the data and the phone itself would make a monitoring graph. The future is already here, even in the most scatological of its versions.