FPGA based system for blood glucose sensing using photoplethysmography and online motion artifact correction using adaline

This paper proposes a non invasive blood glucose sensing system using photoplethysmography (PPG). Neural network based adaptive noise cancellation (adaline) is employed to reduce the motion artifact. Also artificial neural network is used to create the predictive model which estimates the glucose levels based on PPG signals. Error in estimating glucose levels came out to be 5.48 mg/dl using ANN on MATLAB. This predictive model created by ANN has been implemented on FPGA. Error in estimating glucose levels by the ANN model implemented on FPGA, came out to be 7.23mg/dl. The results have been validated by performing Clarke error grid analysis.