Healthcare Technology Featured Article

August 13, 2015

Watson Steps Up Its Medical Chops with X-Ray, MRI Scanning Data

While Watson will forever be remembered as the machine that thoroughly trounced two of the best Jeopardy players to ever play the game—Ken Jennings and Brad Rutter—its use on other fronts is also rapidly coming into play. Some have wondered if it might not be a force in the sales division. Others have wondered if Watson might make a noteworthy addition to the medical field. That last one has left some pondering, especially in light of news that IBM has given Watson new ability to interpret both X-ray results and magnetic resonance imaging (MRI) scans.

Image via IEEE

Last week, IBM announced plans to acquire Merge Healthcare later this year, a deal set to be concluded for the sum of $1 billion. With that acquisition, reports suggest that IBM is planning to train Watson on how to handle the aforementioned imaging tools. With this data in place, reports suggest, Watson might be better able to scan the human body for signs of heart disease or even cancer.

The use of artificial intelligence in recognizing people and objects isn't exactly new—reports suggest that Facebook and Google alike have already been engaged in such studies for a while, using deep learning practices to teach through huge numbers of examples—but the idea that artificial intelligences could use deep learning methods to spot tumors is novel. IBM, via its blog, noted that this acquisition would give the Watson Health platform the ability to advance “...beyond natural language,” effectively “ it the ability to 'see'.” It’s even been suggested that Watson might have a hand in the development of personalized therapies based on the individual genetic makeup of the tumor in question.

Naturally, there are some who think that this may not come about as expected. Johns Hopkins University's John Eng, an associate professor of radiology, notes that there's a lot of “ambiguity and fuzziness” surrounding medical data, and thus a “general diagnostic machine” is still some ways off. However, it's worth noting that without steps like this, the “general diagnostic machine” Eng posits will never exist.

Indeed, that's what it's really about in the end: the first steps. The great advantage of machines is that they can take in more information in minutes than a human could in a lifetime. But the problem is applying that information. This is usually where a human comes in, but if a machine could be taught to apply that information in the same way a human could, it would make the machine a truly amazing tool. Such developments must start somewhere, and when it comes to Watson, this may be the perfect starting point. If something like this were then coupled with telemedicine standards as well as videoconferencing, we may well be able to remove an entire layer from the medical industry as we know it, subcontracted to machines. If doctors were left to only treat patients—if recovery was done at home using conferencing tools, and diagnostics were done by machines—the overall size of the medical industry would drop, and the expense involved would likely go with it.

Watson seems to be well on its way to being one of the great medical tools of our era. It's going to take plenty of development to get it to that point, of course, but the farther it goes, the better off we all are, medically.

Edited by Dominick Sorrentino

FREE eNewsletter

Click here to receive your targeted Healthcare Technology Community eNewsletter.
[Subscribe Now]