(Image CreditL Kalhh/pixabay)
Nowadays, machines can replicate a human's eyesight and taste but haven't caught up on the smell side. Of course, we've seen e-noses that detect cancer in blood cells and examine the air surrounding wastewater treatment plants. However, humans have 400 olfactory receptors, which may explain why AI couldn't mimic the sense of smell.
Researchers from the University of Pennsylvania's Monell Chemical Senses Center and Osmo created an AI model with better odor identification capabilities than humans. The system also analyzed scent molecules and described them like a person. This model eventually paved the way for a Principal Odor Map (POM). The team says this could lead to better mosquito repellants or deodorizing products.
"In olfaction research, however, the question of what physical properties make an airborne molecule smell the way it does to the brain has remained an enigma," said senior co-author Joel Mainland, PhD, Monell Center Member. "But if a computer can discern the relationship between how molecules are shaped and how we ultimately perceive their odors, scientists could use that knowledge to advance the understanding of how our brains and noses work together."
The researchers trained the AI system by feeding it the molecular structure of 5,000 odorants with words to describe certain odors, including "musty" or "minty." Additionally, 15 panelists put their noses to work with 400 odors, describing each one using 55 assigned words. They also rated the term that suited the odor on a 1-to-5 scale. One panelist rated the uncharacterized odorant 2,3-dihydrobenzofuran-5-carboxaldehyde as very powdery (5) and somewhat sweet (3). Ultimately, the AI model outperformed the panelists but achieved a spectacular result.
"The most surprising result, however, is that the model succeeded at olfactory tasks it was not trained to do," said Mainland. "The eye-opener was that we never trained it to learn odor strength, but it could nonetheless make accurate predictions."
Afterward, the team mapped 500,000 odor molecules that hadn't been synthesized before, which would take humans 70 years to achieve. Their model identified dozens of pairs of structurally dissimilar molecules with "counter-intuitively similar smells, and characterize a wide variety of odor properties, such as odor strength, for 500,000 potential scent molecules."
"We hope this map will be useful to researchers in chemistry, olfactory neuroscience, and psychophysics as a new tool for investigating the nature of olfactory sensation," said Mainland.
Have a story tip? Message me at: http://twitter.com/Cabe_Atwell