A world-leader in endoscopic technology, i-Nside has one mission: “to provide healthcare by developing disrupting tools and applied artificial intelligence (AI) software that will fundamentally increase the access to diagnostics in remote and rural areas around the world.” The importance of this goal becomes clear when you look at the stats.
While there is a global shortage of health professionals, this deficit is most pronounced in rural areas. According to the World Bank, over 40% of the world’s population still resides in rural areas. Despite this, healthcare, especially specialized medicine, remains a largely urban affair. This disparity isn’t limited to so-called “developing” countries either. A 2017 report by the National Rural Health Association noted, for instance, that here in the U.S., while there were 263 medical specialists per 100,000 people in urban areas, the ratio in rural communities was just 30 to 100,000.
While medical devices can give primary-care rural health professionals the tools to examine their patients more thoroughly, getting a definitive diagnosis may still require the expertise of an ear, nose, and throat (ENT) specialist. Referrals could see patients traveling for miles for what could end up being a simple procedure. Ear infections, for instance, can be difficult to diagnose, but easy to treat. If a patient can get an accurate diagnosis at the time of their hometown appointment, relief could be as much as a few eardrops away.
As such, i-Nside wanted to build an affordable diagnostic platform that could help these general medical practitioners diagnose inner-ear diseases on the spot, using computer vision. For a small team like theirs, building an AI platform on their own would have been not only challenging but costly. As such, they sought an AI partner, choosing Clarifai over tech giants like IBM Watson for our customer support and “a simple, well-documented API” which allowed them to build their platform without incurring much, if any, “financial or technical debt."
.@i_NsidePro uses machine learning on-device to allow doctors to process images and check for disease without potential privacy violations, the need for a data center or internet connection. https://t.co/i39QJL3d3b pic.twitter.com/ISDheQajb3— Clarifai (@clarifai) 21 June 2018
Providing us with over 100,000 ear images collected from their widely distributed endoscopic tool, Clarifai was able to custom train computer vision model for ear pictures and video. The model was able to detect 10 common inner-ear conditions with over 98% accuracy. They then built an affordable, easy-to-use smartphone attachment, the Smart Scope adaptor, that could be used to quickly examine and take photos of the inner-ear.
Using our iOS EDGE SDK to deploy their custom model in their assisted diagnostics tool, community care health professionals can now use the Smart Scope adaptor to detect and diagnose inner-ear conditions with almost total accuracy, even offline.
As medical professionals worldwide use the Smart Scope and diagnostics app to diagnose inner-ear conditions, machine learning allows i-Nside’s custom model to improve in accuracy. Once each healthcare worker gets back online via the internet or their local data network, their app will reconnect to the Clarifai Cloud, adding to and gaining from that aggregated learning, allowing patients from any corner of the world to benefit from the experience of community doctors and nurses the world over.
From the company’s tagline, we know i-Nside wants “to provide healthcare for everyone, everywhere.” And now, with Clarifai’s EDGE SDKs, they are steps closer to achieving this purpose.