Personalized Goals for Diabetes Self-Management

The goal of this project was to design and develop an automated system that generates personalized nutritional goal recommendations for individuals with type 2 diabetes (T2D). GlucoGoalie takes individual self-monitoring data, including the nutrient composition of meals and blood glucose measurements, and uses machine learning and a rule-based expert system to generate nutritional goals that suggest changes to the macronutrient composition of meals. While any current applications focus on supporting reflection with data, this approach aims to help patients in interpreting and action on their self-tracking data.

Papers and Talks

Mitchell EG, Burgermaster M, Heitkemper E, Levine M, Miao Y, Tabak E, Albers D, Smaldone A, and Mamykina L. 2019. Feasibility of a machine learning based method to generate personalized nutrition goals for diabetes self-management. AMIA Annual Symposium, November 2019. Washington, DC. (Podium Abstract)

Mitchell EG, Levine M, Mamykina L, Tabak E, and Albers D. Machine learning for personalized decision support with patient-generated health data. AMIA Annual Symposium, November 2019. Washington, DC. (Poster)

Mitchell EG, Burgermaster M, Heitkemper E, Levine M, Miao Y, Desai P, Hwang M, Tabak E, Albers D, Smaldone A, and Mamykina L. 2019. Personalized, data-driven recommendations for diabetes self-management with GlucoGoalie. Workshop on Interactive Systems in Healthcare (WISH), May 2019. Glasgow, UK. (Presentation)

© 2018 by Elliot G Mitchell. Proudly created with Wix.com

Last Updated August 2018

  • email
  • google-scholar
  • LinkedIn - Black Circle
  • GitHub