Numerous studies have shown that trained dogs can detect many types of diseases – including lung, breast, ovarian, bladder and prostate cancer and possibly Covid-19 – simply by smell. In some cases, involving prostate cancer, for example, dogs had a 99% success rate in detecting the disease by sniffing patients’ urine samples.
But it takes time to train these dogs and their availability and time are limited. Scientists are looking for ways to automate the incredible olfactory capabilities of the canine nose and brain in a compact device. Now, a team of researchers from MIT and other institutions have created a system that can detect the chemical and microbial content of an air sample with an even greater sensitivity than a dog’s nose. They coupled this with a machine learning process that can identify the distinct characteristics of disease-carrying samples.
The findings, which researchers say may someday lead to an automated odor detection system, small enough to be incorporated into a cell phone, are being published today in the newspaper. PLOS One, in an article by Claire Guest of Medical Detection Dogs in the UK, research scientist Andreas Mershin from MIT and 18 others from Johns Hopkins University, the Prostate Cancer Foundation and several other universities and organizations.
“Dogs, for now 15 years or more, have proven to be the first and most accurate disease detectors for anything we’ve tried,” says Mershin. And his performance in controlled tests in some cases exceeded that of today’s best laboratory tests, he says. “So far, many different types of cancer have been detected earlier by dogs than any other technology.”
Furthermore, dogs apparently pick up connections that have so far eluded human researchers: when trained to respond to samples from patients with one type of cancer, some dogs identified several other types of cancer – although the similarities between the samples were not it is evident to humans.
These dogs can identify “cancers that don’t have identical biomolecular signatures in common, nothing in the odorants,” says Mershin. Using powerful analytical tools, including gas chromatography mass spectrometry (GCMS) and microbial profile, “if you look at samples of, say, skin and bladder cancer and breast and lung cancer – all the things the dog demonstrated be able to detect – they have nothing in common. ”Still, the dog can somehow generalize from one type of cancer to be able to identify others.
Mershin and the team in recent years have developed and continued to improve a miniaturized detector system that incorporates stabilized mammalian olfactory receptors to act as sensors, whose data flows can be handled in real time by the features of a typical smartphone. He imagines the day when each phone will have a built-in odor detector, just as cameras are now ubiquitous on phones. These detectors, equipped with advanced algorithms developed through machine learning, can detect early signs of disease much earlier than typical screening regimes, he says – and may even alert you to smoke or a gas leak.
In the latest tests, the team tested 50 urine samples from confirmed cases of prostate cancer and controls known to be free of the disease, using dogs trained and handled by medical detection dogs in the UK and the miniaturized detection system. They then applied a machine learning program to discover any similarities and differences between the samples that could help the sensor-based system to identify the disease. When testing the same samples, the artificial system was able to match the dogs’ success rates, with both methods scoring more than 70 percent.
The miniaturized detection system, says Mershin, is actually 200 times more sensitive than a dog’s nose in terms of being able to detect and identify tiny traces of different molecules, as confirmed through controlled tests required by DARPA. But in terms of the interpretation of these molecules, “it is 100% dumber”. This is where machine learning comes in, to try to find the evasive patterns that dogs can infer from smell, but humans have not been able to understand with a chemical analysis.
“Dogs don’t know chemistry,” says Mershin. “They don’t see a list of molecules appearing in their heads. When you smell a cup of coffee, you don’t see a list of names and concentrations, you feel an integrated sensation. This sense of olfactory character is what dogs can mine.
Although the physical apparatus for detecting and analyzing molecules in the air has been under development for several years, with a lot of focus on reducing their size, until now there was a lack of analysis. “We knew that sensors are already better than dogs can do in terms of detection limits, but what we haven’t shown before is that we can train artificial intelligence to imitate dogs,” he says. “And now we show that we can do that. We show that what the dog does can be replicated to some extent. “
This achievement, the researchers say, provides a solid framework for future research to develop the technology at a level suitable for clinical use. Mershin hopes to be able to test a much larger set of samples, perhaps 5,000, to pinpoint significant indicators of disease in more detail. But these tests are not cheap: it costs about $ 1,000 per sample for clinically tested and certified samples of urine with and without diseases to be collected, documented, sent and analyzed, he says.
Reflecting on how he got involved in this research, Mershin recalled a bladder cancer detection study, in which a dog mistakenly identified a member of the control group as being positive for the disease, although he had been specifically selected based on hospital tests. how to be disease free. The patient, who knew about the dog test, chose to do more tests and, a few months later, it turned out that he had the disease at a very early stage. “Even though it’s just a case, I have to admit that it influenced me,” says Mershin.
The team included researchers from MIT, Johns Hopkins University in Maryland, Medical Detection Dogs in Milton Keynes, UK, Cambridge Polymer Group, Prostate Cancer Foundation, University of Texas at El Paso, Imagination Engines and Harvard University. The research was supported by the Prostate Cancer Foundation, the National Cancer Institute and the National Institutes of Health.