There are over 450 millions of diabetic people all over the world. More shocking 1 out of 3 people don't even know they have diabetes. Diabetes either type 1 or 2 if not controlled can lead to serious health problems including blindness, heart disease, nerve damage, organ amputation. Controlling diabetes requires often monitoring of blood glucose level which can be either done at hospital or with personal glucometer kit. In both cases blood sample needs to be drawn from the body. Doing glucose test several times a day is obviously painful, moreover more check your glucose better you can control and continuous monitoring of glucose can lead to maximum flexibility to control it. This project aims at open source non-invasive glucometer where people no longer need to draw blood samples in order to measure it. Many attempts have been made to make it happen but none of those came out as a real product. In this approach I'm trying Machine Learning (Artificial Intelligence) and a new sensor to make it happen, currently it's in development stage.
My solution includes a custom hardware packed with new sensor and an embedded processor and a firmware software. Machine Learning algorithm is being optimized so that it can run smoothly in the embedded processor.