Mother creates a machine learning system to monitor her son’s diabetes

Type 1 diabetes (T1D) is a devastating chronic disease that is estimated to affect 80 000 new children each year. This disease is characterised by an impairment on the production of insulin by the pancreas, resulting in high levels of blood glucose levels in the body.

When Vivienne’s son, Felix, was diagnosed with T1D, she and her wife, Norma, were confronted with the blood glucose levels monitoring reality. More than monitoring Felix’s glucose levels, the two scientist started recording everything, including food intake, activity levels, emotional states and every blood glucose levels. In addition to this, they even built a small cyborg containing Fitbits, Basis Bands, wireless insulin pumps and continuous glucose monitors. If these wasn’t enough, they also had to hack the data captured by the wearable devices as it was not accessible. When Vivienne and Norma took all these information to the doctors, the reaction was not what they expected.

“It was truly crazy that my son was wearing an entire wardrobe of sophisticated technology, yet his pump would be set by hand once every few months based largely on intuition.”, said Vivienne.

Immediately, Vivienne began building an algorithm that evolved into Jitterbug, a new machine learning system, that was able not only to monitor but also to predict Felix’s glucose levels. As she wasn’t able to commercialize it, she donated all the codes and data to the TidePool Foundation, a nonprofit building open source code for diabetes.

“Every parent in the world should have that one moment where they feel like they have uniquely added something profound to the story of their child’s life.”, Vivienne said.

Adapted from: http://quantifiedself.com/2018/08/women-taking-control-vivienne-ming/

More info: https://su.org/about/faculty/vivienne-ming/
https://vimeo.com/81272562