The IFTF Blog
The Fascinating Future of Cloud Diagnosis
Via a slightly old article in Good Magazine comes word of a great student project out of Australia called StethoCloud that is aimed at using the receiver on a phone to diagnose pneumonia by capturing the breathing of a potential pneumonia patient and sending that information off to the cloud for analysis. The idea is simple but brilliant: Pneumonia is often treatable with antibiotics, but becomes lethal when diagnosis and treatment take too much time, and so speeding diagnosis could substantially reduce mortality.
Here's how Good describes the process:
The mic captures the sounds of the person breathing and the app uploads the recording onto cloud servers. Then the app analyzes the breathing patterns, makes a diagnosis according to the standards of the World Health Organization—either the subject has pneumonia or doesn't—and then presents the user with the appropriate treatment plan.
While a regular digital stethoscope runs over $600, the StethoCloud only costs about $20, which is significantly more affordable in the developing nations that are home to 98 percent of childhood pneumonia deaths. And, although a phone is required for the system to work, about "1.5 million pneumonia deaths occur in developing countries with a high enough mobile usage that we can directly address it without distributing anything else," says Lin.
The concept reminded me of something called the Parkinson's Voice Initiative which has developed an algorithm that can diagnose Parkinson's Disease by running the subtle details of someone's voice through an algorithm. As Wired notes, the project's director Max Little has a big vision for the algorithm: "If we could use the entire existing telephone network then we could scale up the screening of Parkinson’s disease to the entire population, and do it at very minimal cost.”
The potential value of each of these concepts is enormous. As Good notes, pneumonia is the biggest killer of children under 5 globally. While early diagnosis for Parkinson's, like most diseases, opens up opportunities to start treatments much sooner, slow the disease down, and enable people to start preparing to deal, emotionally and socially, with the disease.
At the same time, it's worth noting that parts of this future feel a bit creepy.
What both of these initiatives point to is a future where an increasingly wide array of illnesses can be diagnosed without people intervening--and, in the case of Parkinson's, the idea is that you don't even have to opt in to being screened; the phone network will simply notice something is amiss, and identify something that your doctor would have no chance of figuring out just by talking to you. In practice, there are some obvious logistical questions: For example, would your cell phone's customer service department give you a call and tell you your voice tremors indicate you need to go see a neurologist? This seems… problematic. Which is to say that we're entering a future where are technical abilities to identify health needs may outpace our social desires for them--a tension that we should expect to increase rapidly over the next decade.