Buoy hopes to fight fake online health news with an artificially intelligent app

Nearly every time I go to WebMD’s symptom checker it somehow leads to telling me I might have some horrible, cancerous disease. Heading over to Yahoo Answers is usually no different. The last time I typed in Google for “shoulder pain” I got an ad recommending I buy a “LightStim” tool promising to shoot “Multi-wave LED technology” through my body.

So when I heard about Buoy, a new app using an intelligent algorithm backed my medical data to guess what ails you I was mildly intrigued. What weird, gross illness will it tell me I have this time?

It told me I may have bursitis. Boring, old bursitis — a condition that often occurs in middle age where the bursa sac in your shoulder gets inflamed from overuse. It made sense, anti-dramatic as it was. I type a lot.

How did Buoy come up with this conclusion? The app works by taking your age, gender, geolocation and other factors into account and then pairing them with millions of medical records based on the symptoms you add. I told the app about my shoulder pain and how long it had been acting up. Tada, probably bursitits, something my doctor, after looking at my shoulder, also told me I likely had and then promptly sent me to an orthopedic specialist for further evaluation.

It turned out it wasn’t bursitis. But both the app and my primary doctor initially concluded the same thing — something, by the way, WebMD also added as an option, along with dislocated shoulder, tendonitis and Gout. That last part is important and should be emphasized with all remote diagnostics tech – it still can’t replace a human.

However, given my age, location and symptoms, it made sense to tell me it was likely bursitis. And, instead of suggesting “one weird trick” to fix it, the app suggested a visit to the doc. Buoy will even suggest doctors and prompt you to make an appointment.

Buoy founder and Harvard-trained MD Andrew Le tells TechCrunch he has implemented “a battery of tests” and conducted “a series of studies” to ensure Buoy’s accuracy — starting with running a quality control trial to compare 500 people’s symptoms to the doctor’s diagnosis. The conclusion? Buoy was right 90 percent of the time, says Le.

Still, the app is not without competition as it seems everyone in healthcare is suddenly diving into AI — with IBM’s Doc Watson as the biggest whale in the space.

Though Le argues Watson uses natural language processing to understand medicine based on medical papers and says that doesn’t translate to the same language patients use to describe their symptoms. Buoy’s differentiator is instead using that same data translated into laymen’s terms so us regulars get an idea of what might be happening in our bodies.

Le’s plan is to onboard both consumers and hospital organizations to swiftly process patients before they see the doctor. Buoy is still in the middle of raising about $3 million in funding from various angel investors but the platform launches today. Those interested in checking out their symptoms can do so on Buoy’s web app and soon iPhone or Android.