Simulation on how to customize glucose adjustment method for non-invasive blood glucose sensing by NIRS

摘要

Previous studies have shown the limitations of taking OGTT (Oral Glucose Tolerance Test) as the glucose adjustment protocol for non-invasive blood glucose sensing. Previous studies built a mathematical model of glucose metabolism system-IMM (the Integrated Minimal Model) to probe other available adjustment methods. In this talk, a further study would be focused on more detailed combination options of different glucose input types for glucose adjustment projects in non-invasive blood glucose sensing. And predictive models of blood glucose concentration have been established by means of partial least squares (PLS) method, which could be used to evaluate the quality of different glucose adjustment options. Results of PLS modeling suggested that predictive models under combined glucose input types, compared with OGTT, show a great enhancement in the stability. This would finally improve the precision of non-invasive blood glucose sensing.