A poster presentation by a team of researchers in the Department of Computer Science and Engineering at Texas A&M University, received runner up for Best Poster Award in the student poster competition at the 2018 Institute of Electrical and Electronics Engineers (IEEE) International Conference on Body Sensor Networks in March.
Dr. Ashutosh Sabharwal, along with Dr. Theodora Chaspari, Bobak Mortazavi, Temiloluwa Prioleau, and Dr. Ricardo Gutierrez-Osuna, won the award for their poster titled “Sparse Representation Models of Continuous Glucose Monitoring Time-series.”
Continuous glucose monitoring (CGM) is essential toward effectively managing Type 1 diabetes. These systems can effectively provide real-time blood-glucose measures and warn individuals regarding dangerously high or low glucose levels. While such systems have a great potential toward improving diabetes related outcomes, the corresponding time-series might contain multiple sources of noise related to sensor limitations, needle drifts and calibration issues. That is why signal processing steps are needed to identify the meaningful signal components and appropriately interpret the underlying information.
In the poster presentation they propose to use sparse representation techniques with appropriately designed dictionaries to express CGM signals as a linear combination of a small set of knowledge-driven atoms. Results on a dataset of 25 patients diagnosed with Type 1 diabetes indicate that the proposed framework is a viable solution for modeling CGM time-series reaching relative reconstruction error of 0.08 and suggest that this approach can be used to interpret the underlying CGM time-series in relation to clinical assessments.
The IEEE Conference on Biomedical and Health Informatics 2018 and the IEEE Conference on Body Sensor Networks 2018, two premier flagship venues in the area of mHealth, health analytics and wearable computers, co-located this year, along with the Health Information Management Systems Society Annual Conference 2018. The joint organization provided a unique forum to showcase novel sensors, systems, signal processing, analytics and data management services.