A team of 56 researchers has designed a Breathalyzer-type device that can accurately detect 17 different diseases. (all images via IT)
The title may read like click-bait, but it couldn’t be further from the truth as 56 researchers from five different countries came together to first prove a theory first by testing a hypothesis made by ancient Greeks and second, to build a device that could help quickly diagnose patients with various diseases.
The idea behind the science is that the chemical makeup of different diseases can be identified through breath samples. Just imagine how the Greeks came up with the idea back in ancient times and what technology was on hand and it doesn’t paint a flattering picture on what the physicians of the day had to endure.
Regardless, turns out the theory was sound, so much so that it’s accurate enough to detect 17 different diseases including kidney cancer, Parkinson’s disease, and even multiple sclerosis. The researchers lead by Professor Hossam Haick of the Israel Institute of Technology, studied more than 1,400 patients from all over the globe that is afflicted with various diseases, collected breath samples and used mass spectrometry to test the chemical composition of each for positive identification.
As a result, thirteen different chemical components in different compounds were found in each of the sample makeups, creating a unique kind of fingerprint for each. They then set out to design a device that could detect those ‘breathprints’ accurately, and as a result, the ‘No-nose’ (otherwise known as the artificially intelligent nanoarray) was born.
The device was designed using two internal nanolayers with one utilizing strictly a carbon layer and the other dusted with modified gold nanoparticles and a grid of nanotubes. While both conduct electricity, the carbon-only layer stores the breath sample or VOC (Volatile Organic Compound), which in turn alters the electrical resistance. It’s by measuring that resistance that produces the unique disease signature.
Table of VOCs mapped by their heat signature and their corresponding diseases.
A unique device obviously needs some unique software to generate results, which is why the researchers turned to AI, not because of the ‘coolness’ factor but because there are literally hundreds of VOCs that are exhaled in each breath. Differentiating between them all is time-consuming, especially when you only need 13 to identify each disease.
The AI adds up all those VOCs in each breath then combs over a database to match the VOCs with their unique breathprint and delivers the diagnosis. To garner the device’s effectiveness, the researchers performed a blind test, meaning they were not aware of which subject had which disease. The results were then measured against the mass spectrometer analysis and were found to be accurate by 87% as to who had what affliction.
The No-nose still needs to undergo development for better accuracy and further testing will be necessary before the device hits the market, but when it does, it will give doctors an inexpensive, fast and easy to use tool to accurately screen and diagnose patients in the near future.
Have a story tip? Message me at: cabe(at)element14(dot)com
Top Comments