Disease-Sniffing Device May Surpass a Dog’s Nose

A new system that can detect the microbial and chemical content of air appears to be more accurate than even the best canine noses and may be a game-changer for the early detection of certain diseases and conditions. For years, doctors and scientists have been working with dogs to sniff out certain types of diseases. But now, science and technology may have advanced to the point that everyone's best friend isn’t needed for this task any longer. A new machine-learning process might be even better at the task. Let’s take a closer look at what we know about this disease-sniffing device so far.

What Diseases Might Dogs Detect?

Doctors and researchers have used dogs to detect certain types of diseases for over 15 years. In fact, during the pandemic, researchers found indications that dogs might even be able to detect COVID-19. Before that, canines were primarily used to detect various types of cancer, including bladder, prostate, lung, breast, and ovarian. When they sniffed men’s urine samples, some dogs showed a 99% success rate in detecting prostate cancer. 

Training dogs to do this takes time, though, making these valuable pups a limited resource not available to everybody. They simply aren’t able to address the needs of every patient who may have these kinds of problems. Yet, because early and accurate detection is key in many cancers, the ability of a dog to sniff out disease at this level is a valuable skill when it’s available. If there were more trained dogs, it might be enough. Fortunately, there’s another option that may make future detection easier.

How Does the New System Work?

The new system for sniffing out disease comes from a collaborative research team from MIT and other locations. The goal of this system is to analyze an air sample for both chemical and microbial agents. It’s even more sensitive than a dog’s nose, so its accuracy should meet or exceed its canine counterparts’ abilities. 

Once the air is analyzed, the machine is then coupled to a machine-learning process that compares it with disease-bearing samples. In other words, it’s about matching up what’s found in the air sample with what would be found in the air sample of someone who had a particular disease. Because it’s based on machine-learning, it’s extremely accurate and highly efficient. That may help reduce both false-positive and false-negative results. 

Is There a Downside?

Right now, the device isn’t yet available for widespread use. It can cost up to $1,000 for the test, and it’s not FDA-approved or covered by insurance. But it shows a great deal of promise for the future and researchers are working on these hangups. The biggest downside is fine-tuning the machine-learning with the analysis of the samples to increase the accuracy. 

When a dog smells something, for example, they don’t understand chemistry. Dogs just know that it smells different from something else. But detecting the why of things isn’t easy, and it’s a fine-tuning issue for the developers of the device. By using knowledge of mammalian olfactory receptors coupled with machine-learning and technology, the device may become smaller, lighter, more accurate, and less expensive over time.

For now, dogs are still better able to detect disease because trained ones recognize specific scents. As the new device becomes “trained” better, it may further surpass what dogs can do and even better, may also be available for widespread use, therefore potentially saving lives.

Copyright 2021, Wellness.com