Massachusetts Institute of Technology
Expressive image
The dog’s nose is a very complex and accurate tool that allows it to distinguish different smells, even to identify the smell of some diseases, such as malaria and certain types of cancer, and also “Covid-19”.
read more
But a dog is a living organism, so it can make mistakes, plus training dogs takes money and time. Therefore, scientists have been seeking for years to combine the accuracy of the “electronic nose” with the ability of an organism to “understand” what the smell is. This appears to have happened recently.
Engineers from the Massachusetts Institute of Technology have devised a system that can determine the microbial and chemical makeup of a sample of air, even better than a dog’s nose. And when machine learning was added to this system, they obtained a device that recognizes the smell of disease in the air.
Massachusetts Institute of Technology
Electronic nose
The researchers wanted to use the dog’s nose, which is characterized by the fact that it feels what a human does not feel, such as the unmatched smells produced by different types of cancer. But trained dogs reveal one of them, and recognize similar smells.
read more
Researcher Andreas Mershin says, “Dogs do not know chemistry, and when they feel the smell, there is no list of molecules in their brain. When a person smells the smell of coffee, he does not have a list of names and ingredients, but only feels a special smell. And such a feeling exists in dogs.”
In order to obtain the same results, the researchers created an automated algorithm that was trained on 50 urine samples from healthy people and others with prostate cancer. This program analyzed the similarities and molecular differences between these samples, and identified with high accuracy the volatile substances associated with the disease.
The results of subsequent tests proved that the effectiveness of the new system in detecting prostate cancer reached 70%, which is no less efficient than dogs.
The researchers indicate, however, that improvements must be made to the new system, and they hope in the future to create a device such as a camera that can be integrated with a smartphone.
Source: Vesti. Ru
