Diabetes Care

Type-1 Diabetes is a chronic illness affecting an estimated 3-million americans. Attentive regulation of blood glucose is needed to avoid dangerous and potentially deadly side effects. Our research aims to offload some of this management work to mobile applications and web-based analytical tools; making use of a combination of contextual computing, continuous glucose monitoring (CGM) technology, and machine learning.

Identifying events of interest from gps traces and activity estimates

tracking arrow clusters

We are building tools that associate daily events with different blood glucose behaviors, and surface these relationships to patients.

Technology for Metabolic Health and Understanding

Type-1 diabetes is an important first environment for learning how to improve blood glucose inference using contextual information, but there are also many potential applications within areas such as type-2 diabetes and athletics.