We developed the technology initially focused on diabetes but we built it modular to easily adapt it to other diseases like epilepsy, high blood pressure and anti-clotting as examples.
Our learning algorithm in the cloud receives information from the third party wearable devices that the patients currently use (continuous glucose monitors and smartwatches), and with that information, it is able to represent on the Smartphone app the predicted future glucose levels and recommended actions to avoid risk (for example, a notification will be sent if the system detects a low glucose risk, with a snack recommendation; or the patient could ask how drinking a coke in 30 minutes or going for a hike will afect their future glucose level). After testing it the error is less tan 10 glucose units when forecasting 1 hour glucose.
With all this information, the system will very soon be able to calculate the optimal insulin dose so we built a working prototype of the insulin smartpatch and designed the functionalities according to the user needs (for example modularity for easy refill, it is comfortable, discrete and safe). We tested it in 2016 and we managed to get drugs through the skin without any kind of needles. We already have an improved prototype to be tested in laboratory during May.