The robot will smell you now: Is odor analysis the next futuristic diagnostic tool?

AI
One expert says AI tools that use odor analysis could be used by clinicians within 3-5 years.

Years from now, machines may be able to diagnose diseases like cancer simply by analyzing a patient’s smell.

Several international companies are using analytics and artificial intelligence to create new diagnostic tools that can pinpoint diseases using odor analysis, according to The New York Times. Developers in England and Israel have begun testing their diagnostic tools in the clinical setting with some success—including one professor who recently published a study showing that his tool could accurately diagnose 17 different conditions based on breath samples.

In the U.S., researchers at the University of Pennsylvania are working to develop a prototype that uses odors in blood plasma to recognize ovarian cancer.

RELATED: 3 ways machine learning will transform healthcare

A study published in January by researchers at Northeastern University indicated that a new class of medical devices known as “e-noses” is gaining traction in the medical community. Cristina Davis, a biomedical engineer and professor at the University of California, Davis, estimated odor analysis tools could be available within the next three-five years.

“I think the fact that you’re seeing so much activity both in commercial and academic settings shows that we’re getting a lot closer,” she told The New York Times.

RELATED: Increasingly powerful AI systems are accompanied by an 'unanswerable' question

Researchers have been exploring the use of machine learning as a diagnostic tool with other sensory outputs, using new technology to detect imperceptible changes in a patient’s voice that could be an indication of post-traumatic stress disorder. A precision medicine initiative at Mount Sinai is using artificial intelligence to classify and predict disease.

However, researchers have also warned of machine learning “voodoo,” while others have pointed out that the better AI systems are at diagnosis, the harder they are to understand.

Suggested Articles

A handful of major healthcare groups and vendors are throwing their weight behind ONC's information blocking proposal.

A recent digital health conference highlighted the ongoing obstacles to using next-generation technology to impact health outcomes.

Hospitals are highly vulnerable to cyberattacks, and the potential for financial disruption is high, according to a Moody's report.