Since I retired I lead a non profit [0] that aims at inventing medical tools like continuous glucose monitoring. We did not made any breakthrough, it is just engineering on paper with little PoC as there is little money. The inventions were published without IP claim, there are one or two downloads each week, but nobody contacted us.

This year we registered on HackaDay[1] for a early heart failure (HF) detector. Most aging adult suffers from HF, sometimes as early as when they are in their forties.

Our design is good, the gold standard here is the Physionet 2016 competition, Physionet 2016 competitors mostly used machine learning on file wide features such as the heart rate and its variability. We choose to find features in heart sounds, there are up to four sounds per beat. The code is on Github. The project on HaD will be declared finished in September or October.

We will have huge problems when we will reach the point of trying to obtain a regulator agreement (EC/FDA). It needs to make travels, provide samples, make clinical studies, rise awareness. Lot of problems that we are not equipped to manage.

If we succeed at this there are other R&D projects in the pipeline, to keep us busy.

[0] https://padiracinnovation.org/

[1] https://hackaday.io/project/19685-early-and-low-cost-detecti...

[2] https://github.com/Hjertesvikt