Activity Recognition

Background

Chronic metabolic diseases such as diabetes are increasingly prevalent in modern society. While traditional interventions focus dominantly on the control of “what to eat”, a recent study on mice found that restricting the timing of eating prevents obesity and metabolic syndrome, suggesting a potential role of “when to eat” in metabolic regulation. This gives hint to the development of new mobile health technologies that support the management and prevention of metabolic diseases such as diabetes by helping users to track and to adjust mealtime rhythm. However, it is often difficult for people to keep tracking the time of each meal from day to day —a phenomena known as tracking fatigue. This project aims to develop algorithms for automatic detection and modelling of mealtime rhythm based on bio-signals that can be readily measured with wearable devices such as activity trackers, continuous glucose monitoring sensors and hearables.

Method

Machine learning

Contributors

• Ms. Lauriane Bertrand (exchange student in 2020 from INP-ENSEEIHT, France)

• Mr. Nathan Cleyet-Marrel (exchange student in 2020 from INP-ENSEEIHT, France)

Selected Publications

• Bertrand L, Cleyet-Marrel N, Liang Z. (2021) Recognizing eating activities in free-living environment using consumer wearable devices. Engineering Proceedings 6(1): 58. Doi:10.3390/I3S2021Dresden-10141.

• Bertrand L, Cleyet-Marrel N, Liang Z. (2021) The role of continuous glucose monitoring (CGM) in automatic detection of eating activities. In Proceedings of the IEEE 3rd Global Conference on Life Sciences and Technologies (LifeTech 2021), Nara, Japan.