How do you smell it?

 In 1960, the release of the movie "Scent of Mystery" marked a bizarre event in film history: this was the first and last time the shameless "Smell-O-Vision" was unveiled-in order to show the audience For dynamic olfactory experience and familiar visual and auditory effects, the theater is equipped with Rube Goldberg-style* equipment that can deliver different smells directly to the seats.

　　*Translator's Note Rube Goldberg is an American cartoonist. He created a mechanism in comics to accomplish some very simple tasks in a tortuous way. This style is used to describe things that are exaggerated, extremely complicated and unnecessary (2021 hermes bags).

　　 However, viewers and critics soon discovered that this experience was terrible. Smell-O-Vision is full of technical problems and has become a dark history (running gag) in the entertainment industry. Even so, the failure of Smell-O-Vision did not prevent entrepreneurs from continuing to pursue the dream of delivering scents to consumers through digital scent technology, especially in recent years.

　　 Similar efforts made headlines, but with little success. Part of the reason is that people's understanding of how the brain converts odorous chemicals into smell is still very limited-this process is still unknown to scientists.

　　 A study by a neurobiologist at Harvard Medical School provides new insights into the mystery of smell. Researchers published an article in the journal Nature on July 1, describing for the first time how the olfactory cortex encodes different odors.

Using carefully selected odors of molecular structures for stimulation and analyzing the neural activity of awake mice, the research team found that the neural representation of odors in the cortex reflects the chemical similarity between odors, so that odors can be classified by the brain. And these representations can be reconnected through sensory experience.

　　 These findings suggest a neurobiological mechanism to explain why people have a common but highly personalized olfactory experience.

"We all have a common olfactory perception framework. You and I agree that lemons and limes smell similar and agree that their smell is different from pizza. But until now, we still don't know how the brain organizes such information. "Sandeep Robert Datta, an associate professor and senior research author from the Blavatnik Institute at HMS, Harvard Medical School, said.

　　 These results open up new research avenues for a better understanding of how the brain converts the chemical information of smells into smell perception.

　　Datta said: "This is the first explanation of how the olfactory cortex encodes the chemical information of smell. The chemical information of smell is the basic sensory cues of smell."

　　 calculate the smell

　　 The sense of smell allows animals to recognize the chemical nature of the world around them. Sensory neurons in the nasal cavity detect odor molecules, and then transmit the signal to the olfactory bulb (replica hermes). The olfactory bulb is the forebrain structure for primary olfactory processing. The olfactory bulb mainly transmits information to the piriform cortex, which is the main structure of the olfactory cortex, which can process odor information more comprehensively.

Unlike light or sound, which are easily controlled by adjusting the frequency and wavelength, it is difficult to explore how the brain builds neural representations of small molecules that transmit odors. Usually, subtle chemical changes, such as the number of carbon atoms here or the number of oxygen atoms there, can cause significant differences in smell.

Datta and the study’s first author Stan Pashkovski, a researcher at Harvard Medical School, and their colleagues, break through this by focusing on how the brain recognizes related but different smells. problem.

　　Datta said: “The fact is that almost everyone thinks that lemons and limes smell similar, which means that their chemical composition must somehow evoke similar or related neural representations in the brain.”

　　 In order to investigate, the researchers developed a method to quantitatively compare odorous chemicals, similar to using wavelength differences to quantitatively compare the color of light.

　　 They used machine learning technology to study the chemical structure of thousands of known odors, and analyzed thousands of different characteristics of each structure, such as atomic number, molecular weight, electrochemical properties, etc. Together, these data allow researchers to systematically calculate how similar or different any odor is relative to another.

The team designed three sets of odors from the database: one set of odors has a high degree of difference; one set has a medium difference, and the odors are divided into related clusters; the other group has low variability, where the odor is only due to the increase in carbon chain length And change.

　　 Then, they exposed the mice to various odor combinations in different groups and used a multiphoton microscope to image the neural activity in the piriform cortex and olfactory bulb.

　　The prediction of smell

　　 Experiments show that the similarity of neural activity reflects the similarity of odor chemistry. The associated odors produced related neuronal patterns in both the piriform cortex and the olfactory bulb, which manifested as an overlap of neuronal activity measured in the two regions. Conversely, weakly related odors produce weakly related activity patterns.

　　 Compared with the olfactory bulb, in the cortex, related odors lead to more concentrated patterns of neural activity. The researchers observed this phenomenon in different mice. Cortical representations of odor relationships are so relevant that measurements made on one mouse can be used to predict the specific recognition of odor in another.

Further analysis determined a series of chemical characteristics, such as molecular weight and certain electrochemical properties. These characteristics are related to the pattern of neural activity. The information gathered from these characteristics is sufficient to predict the cortical response of another animal to another set of odors based on the results of experiments performed on one set of odors.

　　 Researchers also found that these neural representations are very flexible. The mice were repeatedly given a mixture of two odors, and over time, the corresponding neural patterns of these odors in the cortex became more closely related. This happens even if the chemical structures of the two odors are different.

The adaptive ability of the cortex is partly produced by a network of neurons that selectively reshape the relationship between odors. When the normal activity of these networks is blocked, the way the cortex encodes smells will be more like the olfactory bulb.

　　Datta said: "We present two odors as if they have the same origin, and observe that the brain can rearrange itself to reflect the passive olfactory experience."

　　 Something like lemons and limes smell similar, partly because animals of the same species have similar genomes and therefore have similar smells. But each individual also has a personalized sense of smell.

　　Datta said: "The plasticity of the cortex may help explain why each individual can perceive odors uniquely, although odors are constant from individual to individual.

　　 The results of this study shed light on how the brain encodes smells for the first time. Compared with the relatively easy-to-understand visual and auditory cortex, how the olfactory cortex converts information about odor chemistry into smell remains unclear.

　　The research team said that understanding how the olfactory cortex maps similar odors now provides new insights for understanding and possibly controlling odor perception.

　　 "We still don't fully understand how chemical substances are transformed into perception." Datta said, "There is no computer algorithm or machine that uses a certain chemical structure to tell us what a chemical substance will smell.

"If we want to really make a machine that one day can create a controllable virtual world of smell for people, we still need to understand how the brain encodes information about odors." Datta said, "I hope our findings are along the lines. A step on this path."