A non-invasive glucose level estimation in a multi-sensing health care monitoring system.

BACKGROUND AND OBJECTIVE: The treatment and care of patients with chronic diseases depends directly on the evolution of biomedical parameters. It is important to have a monitoring health care system that provides biomedical data at any time and place. Here, a multi-sensing health care monitoring system with a built-in non-invasive blood glucose level estimation method is presented.

METHODS: Six biomedical parameters were obtained from 15 participants. Glucose levels were obtained using a computer vision approach. A standard glucose laboratory test was taken as a baseline, and a commercial glucometer as a secondary reference. The remaining parameters were also contrasted with a commercial vital signs monitor.

RESULTS: In comparison to standard test, our proposal reported a better performance (RMSE of 9.811) than obtained with the commercial glucometer; the Mann-Whitney test found no significant differences. The remaining biomedical parameters exhibit similar results to the commercial vital signs monitor as validated by a cardiologist.

CONCLUSION: The results suggest the proposed approach could be considered highly competitive regarding standard tests and validated with commercial health care monitoring systems.