Numerous studies have shown that trained dogs can detect many types of diseases – including lung, breast, ovarian, bladder and prostate cancers and possibly Covid-19 – simply by smell. In some cases, for example with prostate cancer, the dogs achieved 99 percent success in detecting the disease by sniffing patients’ urine samples.
But it takes time to train such dogs, and their availability and time are limited. Scientists have been looking for ways to automate the amazing olfactory ability of the dog’s nose and brain, in a compact device. Now a team of researchers at MIT and other institutions have devised a system that can detect the chemical and microbial content of an air sample with even greater sensitivity than a dog’s nose. They linked it to a machine learning process that can identify the characteristic features of the disease-carrying specimens.
The findings, which researchers say could one day lead to an automated odor detection system small enough to be recorded in a cell phone, are being published in the journal today. PLOS Een, in a paper by Claire Guest of Medical Detection Dogs in the UK, research scientist Andreas Mershin of MIT, and 18 others at Johns Hopkins University, the Prostate Cancer Foundation and several other universities and organizations.
“Dogs have been shown, for about 15 years, to be the earliest, most accurate disease detectors for anything we’ve ever tried,” says Mershin. And their performance in controlled tests has in some cases surpassed the best of current laboratory tests, he says. “So far, many different types of cancer have been detected by dogs in the past than any other technology.”
What’s more, the dogs apparently take up compounds that human researchers have evaded so far: When trained to respond to samples from patients with one type of cancer, some dogs then identified several other types of cancer – although the similarities between the samples was not. is not clear to people.
These dogs can “identify cancers that do not share the same biomolecular signatures, nothing in the fragrances”, says Mershin. Using powerful analytical tools, including gas chromatography mass spectrometry (GCMS) and microbial profiling, “if you analyze the samples of, say, skin cancer and bladder cancer and breast cancer and lung cancer – anything the dog can prove to detect – they have nothing in common. Yet the dog can somehow generalize from one type of cancer to be able to identify the other.
Over the past few years, Mershin and the team have developed a miniaturized detector system that contains the odor receptors of mammals, stabilized to act as sensors, the data streams of which can be handled in real time by the capabilities of a typical smartphone. He foresees a day when an odor detector will be built into every phone, just as cameras are currently ubiquitous in phones. Such detectors, equipped with advanced algorithms developed by machine learning, can pick up early disease signs much faster than typical screening regimes, and he can even warn against smoke or a gas leak.
In the latest tests, the team tested 50 samples of urine from confirmed cases of prostate cancer and controls known to be disease-free, using both dogs trained and handled by Medical Detection Dogs in the UK and the miniature detection system. They then applied a machine learning program to eliminate the similarities and differences between the samples that the sensor-based system could help identify the disease. By testing the same samples, the artificial system was able to match the success rates of the dogs, with both methods achieving more than 70 percent.
The miniaturized detection system, Mershin says, is actually 200 times more sensitive than a dog’s nose in terms of being able to detect and identify small traces of different molecules, as confirmed by controlled tests required by DARPA. But in terms of the interpretation of the molecules, ‘it’s 100 percent stupider’. This is where machine learning comes in, trying to find the elusive patterns that dogs can deduce from the smell, but man could not yet understand a chemical analysis.
“The dogs know no chemistry,” Mershin says. “They do not see a list of molecules appearing in their head. When you smell a cup of coffee, you do not see a list of names and concentrations, you feel an integrated sensation. The sense of smell character is what the dogs can exploit. ”
While the physical apparatus for detecting and analyzing the molecules in the air has been developed for several years, with much of the focus on reducing its size, analysis has so far been lacking. “We knew that the sensors were already better than the dogs could do in terms of the detection limit, but what we have not shown before is that we can train an artificial intelligence to mimic the dogs,” says he. And now we have shown that we can do it. We have shown that what the dog does can be repeated to some extent. ‘
According to these researchers, this achievement provides a solid framework for further research to develop the technology to a level suitable for clinical use. Mershin hopes to be able to test a much larger set of samples, perhaps 5,000, to determine the important indications of diseases in greater detail. But such tests do not come cheap: it costs about $ 1,000 per sample for clinically tested and certified samples of disease-carrying and disease-free urine to collect, document, ship and analyze, he said.
Mershin reflected on how he became involved in this research, recalling a study of the detection of bladder cancer, in which a dog incorrectly identifies one of the members of the control group as positive for the disease, although he specifically based on hospital tests chosen to be disease free. The patient, who knew of the dog’s test, chose to do further tests, and a few months later it was found that he had the disease very early. “Even though it’s just one case, I have to admit it affected me,” Mershin says.
The team included researchers at MIT, Johns Hopkins University in Maryland, medical detection dogs in Milton Keynes, UK, the Cambridge Polymer Group, the Prostate Cancer Foundation, the 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.