Starting way before royal physicians smelled the stools and urine of infant noblemen, doctors have used their sense of smell to detect disease. According to an article in Scientific American, physicians have “known that the breath contains clues to many diseases” since 400 B.C. In the journal Biomedical Chromatography, Buszewski and others wrote, “It has been demonstrated that modern analytical instruments allow the determination of many compounds found in human breath both in normal and anomalous concentrations.”
Researchers at the Israel Institute of Technology reported in a study published by the American Chemical Society that they were successful in diagnosing and classifying 17 diseases from 1404 subjects via pattern recognition analysis of exhaled molecules with a single breathalyzer. In blind testing experiments, the researchers achieved 86 percent accuracy using an artificially intelligent nanoarray of molecularly modified gold nanoparticles. In tests validated by gas chromatography linked with mass spectrometry, they found 13 exhaled chemical species, called volatile organic compounds, associated with the 17 specific diseases, differentiated by the assembly of the compounds. The use of artificial intelligence to analyze big data via a preprogrammed particle array in a breathalyzer is an astonishing breakthrough.
The implication of a single, inexpensive breathalyzer to screen, diagnose, and follow up with a number of diseases teases the potential of further developments. If accurate pattern analysis recognition is possible for an ever-growing number of diseases, applications in home, clinic, and remote field settings can vastly expand and improve the reach of medical care.