AI and diabetes

 It will be nice to bind together, food intake, CGM measurements and AI models to optimize intake for better glucose profiles for diabetes type I and II.

There are pretrained AI models for diabetes.

GlucoNet Model: Deep learning model that can predict Glucose levels based on carbohydrate intake.

Deep Gluco Model: Not only takes carbo intake into consideration but also insulin dosage, physical activity, etc.

GlucoNet is adequate for diabetes type II. I will be homing in on type II.

These are used to improve the MIT control loop for adjusting optimal carbo intake.

There is company (Alertgy) in Melbourne, Australia involved in the development of diabetes monitored by a wearable device.

Right now, I have one with questionable activity, but qualitatively measuring (Zero, when not connected and 3 peaks to coincide in time with break fast, lunch and dinner).

Big data to aid Big apple

Big data shakes hands with academics.

Big cities are sources of big data. Data comes in streams from all directions from diverse sources like traffic, crime, scams, events, emails, social media sites, weather sensors, sniffers of various kinds, parking lots and you name it. This data is not the kind of data that you can feed directly to a relational engine, it is diverse, a mélange and can only be termed informatics.

A new center, Center for Urban Science and Progress (CUSP) at New York University will be utilizing Big data to understand the dynamics of the city through various sensors and look through the data they bring in. Similar programs are already in place in University of Chicago and MIT. Outside USA similar efforts are in place like Spain and Singapore.

NYU will offer MS and PhD programs and takes an integrated look at physical, biological and informational data form the city's sensors. The data comes from not only installed sensors to measure temperature, pollution etc.  but also from body worn sensors like FitBit, and Up wristbands that volunteers wear and go round the city. Of course data from twitter feeds, blogs, video cameras etc. will be integrated as well.

Well there is no dearth for data from such a large resource as Big Apple.

The program has sponsors committed to Big Data with biggies like Microsoft, IBM, Xerox, City agencies, MTA and a couple of National Laboratories thrown in for good measure.

This post was written after reading this article, ‘Smart Cities will need big data’ in Physics today a journal published by the American Physical Society, September 2013, pp. 19, 20. For details get a copy of the journal from a library near you.

Related resources (Please do not skip):

http://cusp.nyu.edu/

Watch this video:

http://cusp.nyu.edu/the-promise-of-urban-informatics-video/

http://www.dw.de/smart-city-santander-takes-lead-in-europe/a-16945649

http://senseable.mit.edu/livesingapore/